U.S. patent application number 13/453691 was filed with the patent office on 2012-10-18 for systems and methods for detecting and using an electrical cochlear response ("ecr") in analyzing operation of a cochlear stimulation system.
This patent application is currently assigned to UNIVERSIDAD AUTONOMA METROPOLITANA. Invention is credited to JUAN MANUEL CORNEJO CRUZ, MARIA DEL PILAR GRANADOS TREJO.
Application Number | 20120265270 13/453691 |
Document ID | / |
Family ID | 41164614 |
Filed Date | 2012-10-18 |
United States Patent
Application |
20120265270 |
Kind Code |
A1 |
CORNEJO CRUZ; JUAN MANUEL ;
et al. |
October 18, 2012 |
SYSTEMS AND METHODS FOR DETECTING AND USING AN ELECTRICAL COCHLEAR
RESPONSE ("ECR") IN ANALYZING OPERATION OF A COCHLEAR STIMULATION
SYSTEM
Abstract
Methods and systems for analyzing operation of a cochlear
stimulation system. A sound stimulus signal is generated to excite
the cochlear stimulation system to operate. During operation, the
intracochlear electrodes generate signals into the auditory nerve
system. The patient's nervous system's response may be measured as
the Electrical Cochlear Response ("ECR"). The ECR can be detected
and analyzed for fitting, calibration, performance evaluation and
failure detection of the cochlear implant of the patient. Also
example methods may be used to estimate the audiometric thresholds
of the cochlear implant without the implanted patient's
knowledge.
Inventors: |
CORNEJO CRUZ; JUAN MANUEL;
(DEL IZTACALCO, MX) ; GRANADOS TREJO; MARIA DEL
PILAR; (DEL IZTACALCO, MX) |
Assignee: |
UNIVERSIDAD AUTONOMA
METROPOLITANA
DELEGACION TLALPAN
MX
|
Family ID: |
41164614 |
Appl. No.: |
13/453691 |
Filed: |
April 23, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12247997 |
Oct 8, 2008 |
8165687 |
|
|
13453691 |
|
|
|
|
12037323 |
Feb 26, 2008 |
8065017 |
|
|
12247997 |
|
|
|
|
60891582 |
Feb 26, 2007 |
|
|
|
Current U.S.
Class: |
607/57 |
Current CPC
Class: |
A61B 5/121 20130101;
A61B 5/04845 20130101; A61N 1/36038 20170801; A61N 1/36039
20170801 |
Class at
Publication: |
607/57 |
International
Class: |
A61F 11/04 20060101
A61F011/04; A61N 1/36 20060101 A61N001/36 |
Claims
1. A system for analyzing operation of a cochlear stimulation
system implanted in a patient, the system comprising: a sound
generating system for generating a sound stimulus signal to elicit
operation of the cochlear stimulation system; an electrical
cochlear response ("ECR") detection system to detect an ECR
waveform in a plurality of electrical signal responses received
from the patient using surface electrodes, the electrical signal
responses being generated in response to the sound stimulus signal,
and the ECR waveform being indicative of operation of the cochlear
stimulation system.
Description
RELATED APPLICATIONS
[0001] This application is a continuation application that claims
the priority of U.S. patent application Ser. No. 12/247,997, titled
"SYSTEMS AND METHODS FOR DETECTING AND USING AN ELECTRICAL COCHLEAR
RESPONSE ("ECR") IN ANALYZING OPERATION OF A COCHLEAR STIMULATION
SYSTEM'' by inventors Juan Manuel Cornejo Cruz and Maria del Pilar
Granados Trejo, which was filed on Oct. 8, 2008. The contents of
U.S. patent application Ser. No. 12/247,997 are incorporated by
reference herein in their entirety.
FIELD OF THE INVENTION
[0002] The present invention relates to cochlear stimulation
systems, and more particularly to methods and systems for obtaining
and using an Electrical Cochlear Response ("ECR") in fitting,
calibrating and evaluating operation of a cochlear stimulation
system.
BACKGROUND OF THE INVENTION
[0003] Hearing impaired individuals typically suffer from a loss of
hearing that falls in one of two general categories: conductive and
sensorineural. Conductive hearing loss results from a failure in
the mechanical chain in the external and middle ear that captures
and drives the sound to the cochlea. Sensorineural hearing loss is
due to the deficiency or damage in the cochlea, particularly of the
hair cells located in the cochlea, which converts the sound to
electrical signals that are transmitted by the auditory nerve to
the part of the brain that creates the sensation of hearing.
[0004] Conductive hearing losses may be corrected, at least
partially by medical or surgical procedures or by using
conventional hearing aids to amplify the sound in order to increase
its energy and patient be able to perceive sounds of the external
word. Sensorineural hearing loss on the other hand may be corrected
using a cochlear stimulation systems or, cochlear implant.
[0005] Cochlear stimulation systems operate by converting sound to
electrical signals, which are applied to the residual auditory
system through an intracochlear electrode array. The intracochlear
electrode array provides electrical stimulation directly to the
auditory nerve fibers to create a sound perception in the brain of
a patient using the cochlear stimulation system.
[0006] A typical cochlear stimulation system includes an audio
pickup, or input (for example, a microphone), an amplifier, a sound
processing system, and a receiver/stimulator coupled to an
intracochlear electrode array. The intracochlear electrode array
and receiver/stimulator are typically part of an implanted portion
of the system. The audio pickup, amplifier and sound processor are
part of the external components of a cochlear stimulation system.
The audio pickup is typically located on an earpiece having a
connection to the sound processing system. The sound processing
system also connects, wirelessly or via a wire, to a transmitter
that is typically attached to the patient's head near the audio
pickup earpiece. The transmitter is attached to the head at a
location that is closest to a receiver connected to the implanted
portion. The transmitter typically communicates with the receiver
via a magnetic coupling. The implanted portion includes electronics
that is coupled to the intracochlear electrodes or intracochlear
electrode array. The intracochlear electrodes extend and terminate
sequentially in a straight or spiraling line. The intracochlear
electrodes are inserted into the cochlear tissue along the
spiraling line that follows the spiral formed by the structure of
the cochlea.
[0007] The intracochlear electrodes are assigned frequency bands in
the auditory frequency range in order from highest frequency bands
to lowest such that the highest frequency band electrodes are
processed closest to the electronics in the implanted portion; the
lowest frequency bands are processed closest to the end of the
spiraling line, near the apex, i.e., the conical tip of the
cochlea. The ordering of the frequency bands conforms to the
functional structure of the cochlea, which is known to process
incoming sound representing the highest frequencies at the base,
i.e., beginning of the cochlea's spiral shape. Low frequencies are
processed by the cochlear tissue extending further into the spiral
shape in descending order, such that the lowest frequencies are
processed near the apex.
[0008] During operation of the cochlear stimulation system, the
audio pickup receives sound input and transmits the electrical
signals to the sound processing system. The sound processing system
multiplexes the signal by filtering the signal at a bank of
bandpass filters connected in parallel. Each bandpass filter in the
bank of bandpass filters corresponds to a different one of the
intracochlear electrodes. The filtered signal is then assigned a
current simulation level, which corresponds to a current of the
signal to be output at the corresponding intracochlear electrode.
The current stimulation level delivered to the cochlea by each
intracochlear electrode is adjusted hopefully according patient's
loudness sensation. Assigning frequency bands and setting a current
level to each intracochlear electrode allows the cochlear
stimulation system to represent incoming sound signal into an
activation sequence to the intracochlear electrodes selected
according to a stimulation strategy programmed into the sound
processing system (described below). Basically, the current
stimulation level is selected from a voltage level or some other
indicator of the sound intensity of the input sound signal.
[0009] The filtered signal at the assigned current stimulation
level are then de-multiplexed and sent to the transmitter. The
transmitter transmits the de-multiplexed signal using a magnetic
coupling to the receiver in the implanted portion. The signal is
multiplexed to extract the filtered signals and each filtered
signal is coupled to the individual intracochlear electrode
corresponding to the filtered signal's bandwidth. The filtered
signals excite the nerve fibers at the location of the
corresponding intracochlear electrodes at a current level that is
intended to correspond to the sound intensity level of the input
sound. The patient senses the sound as the combination of
frequencies corresponding to the intracochlear electrodes that
generated the filtered signal and the combination of sound
intensities corresponding to the current levels at each
intracochlear electrode.
[0010] When a patient is provided with a cochlear stimulation
system, a surgical procedure is performed to implant the components
referred to above as being part of the implanted portion inside the
ear. During the procedure, the intracochlear electrodes are
inserted into the cochlea, and the receiver is implanted in an area
of the ear that is opposite a space where the transmitter may be
placed. The patient is also provided with the transmitter and audio
input connected to the sound processing system.
[0011] A few weeks after the implant procedure, the cochlear
stimulation system is also "fitted" for operation. The purpose of
fitting the cochlear stimulation system is to adjust the range of
current stimulation levels for each intracochlear electrode. The
adjustment is necessary to ensure that the minimum current
stimulation levels correspond to the lowest possible threshold
sound intensity level that the patient can hear, and a maximum
current stimulation level that will not result in pain or
discomfort at high sound levels. That is, fitting permits a
physician to determine the minimum and maximum psychophysical
values of the stimulation current for each intracochlear
electrode.
[0012] Cochlear stimulation systems are programmed to use a minimum
and a maximum current value that hopefully match the hearing
threshold level and most comfortable loudness level of the patient.
The current stimulation level typically refers to a minimum and a
maximum value, depending on the specific cochlear stimulation
system, i.e. the electric current dynamic range. The full range of
current stimulation levels corresponds to a range of sound pressure
levels (in dB.sub.HL) mapped according to the loudness perception
of the patient. The sequence or order of activation of the
intracochlear electrodes depends on the input sound features and
stimulation strategy selected by the clinician, i.e., the code used
to activate a subset of intracochlear electrodes according to the
most important features of the incoming sound. The fitting of the
implant involves generating a "MAP" of ranges of intracochlear
electrode current stimulation levels, preferably meeting the
particular needs of the patient. This means setting a threshold
current stimulation level (or T level) and a maximum comfort level
(or C level) for each electrode. Cochlear stimulation systems
typically provide a procedure that allows a physician to set a T
level and C level as well, to a desired value. It is assumed for
purposes of this disclosure that the cochlear stimulation system
being fitted provides such a facility, either using a manual mode
that may be driven by software, or an automatic mode that permits
downloading the T level from a computer or some other electronic
device.
[0013] A variety of strategies exist for determining the T levels
for each intracochlear electrode in a cochlear stimulation system.
In some cases, the physician may choose to leave the cochlear
stimulation system set to the T levels set by the manufacturer or
use T levels in preconfigured maps of T levels to sound levels. The
values of psychophysical parameters such as current stimulation
levels are highly dependent on the physiology of the patient.
Therefore, it is unlikely that predefined current stimulation
levels would be suitable for many patients.
[0014] The physician may also use a subjective method where the
physician stimulates the patient using a low level electrical
current and increases the electric current level until the patient
informs the physician that he can `hear` the sound. The subjective
method, however, cannot be implemented with children that cannot
yet communicate. In fact, it is likely that any patient cannot
communicate if they are experiencing the sense of hearing for the
first time. Moreover, the patient is typically sedated from the
implant procedure, which requires at least waiting until the
patient can communicate in some way to perform the fitting.
[0015] Objective fitting methods have been developed for use with
the patient sedated and possibly with children as well. Present
objective fitting techniques measure physiological responses, such
as the evoked compound action potential (ECAP), the middle ear
reflex (MER), and the stapedius reflex (SR), to direct electrical
stimulation of the intracochlear electrode. Cochlear stimulation
systems that use objective fitting techniques typically include
hardware and software components that provide the physician with
control over the intensity of the electrical signals applied
directly to the intracochlear electrodes. The electrical signals
are typically biphasic, amplitude balanced pulses generated by an
electrical signal source that is external to the cochlear
stimulation system. The physiological responses are measured using
either surface electrodes such as electroencephalographic ("EEG")
electrodes, cochlear stimulation system intracochlear electrodes
themselves or implanted electrodes, and the objective is to measure
the response of the auditory nervous system to the applied
electrical signals.
[0016] Known objective fitting techniques suffer from various
drawbacks. First, such methods typically require the use of special
fitting components that are part of the cochlear stimulation
systems. The special fitting components are often proprietary
apparatuses and methods designed for exclusive use with particular
cochlear stimulation systems. Second, the techniques require
generating electrical stimulation to the intracochlear electrodes
that bypass the operation mode of the sound processing system of
the cochlear stimulation system. Third, the techniques generally
proceed by setting a T level for some of the intracochlear
electrodes one at a time. This is time-consuming when setting the T
level for all of the intracochlear electrodes and not very accurate
when extrapolating from the T levels determined for a set of
intracochlear electrodes to determine T levels for the rest.
Fourth, the fitting does not factor in sound at all. The
physiological response is a response to an electrical signal, and
not sounds.
[0017] Known objective fitting techniques have been determined to
result in a poor correlation between the threshold levels indicated
by psychophysical measurements, for example, and T and C levels. In
many cases, techniques that rely on direct stimulation to measure
ECAP, MER, SR, and other physiological responses typically result
in an overstimulation of the intracochlear electrodes during
operation. These known objective techniques work by measuring
responses to stimulation of single electrodes. This approach does
not factor in that the physiological responses are different when
processing actual sounds that involve the cumulative effect of
multiple electrodes.
[0018] In view of the above, there is a need for improved systems
and methods for performing objective fitting of cochlear
stimulation systems.
SUMMARY OF THE INVENTION
[0019] In view of the above, improved systems and methods for
fitting, calibrating, and/or otherwise analyzing operation of a
cochlear stimulation system are provided. In one aspect of the
invention, an example of a system is provided for analyzing
operation of a cochlear stimulation system implanted in a patient.
The system includes a sound generating system for generating a
sound stimulus signal to elicit operation of the cochlear
stimulation system. An electrical cochlear response ("ECR")
detection system processes a plurality of electrical signal
responses received from the patient using surface electrodes to
detect an ECR waveform. The electrical signal responses being
generated in response to the sound stimulus signal. The ECR
waveform being indicative of operation of the cochlear stimulation
system.
[0020] In another aspect of the invention, an example method is
provided for analyzing operation of a cochlear stimulation system
implanted in a patient. According to the example method, a sound
stimulus signal having at least one selected frequency and sound
intensity is generated. A plurality of electrical signal responses
is generated in response to the sound stimulus signal. The
electrical signal responses are processed as measured responses to
the sound stimulus signal at generated frequencies and sound
intensities. The measured responses are analyzed to determine if
the electrical signal responses include an electrical cochlear
response ("ECR") waveform. The ECR waveform being indicative of
operation of the cochlear stimulation system.
[0021] Other systems, methods and features of the invention will be
or will become apparent to one with skill in the art upon
examination of the following figures and detailed description. It
is intended that all such additional systems, methods, features and
advantages be included within this description, be within the scope
of the invention, and be protected by the accompanying claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0022] The above and other objects of this invention will become
more readily apparent upon reading the following text and drawings,
in which:
[0023] FIG. 1A is a graph depicting an example ECR waveform.
[0024] FIG. 1B shows a set of signals illustrating operation of a
system for obtaining an ECR waveform.
[0025] FIG. 1C is a flowchart illustrating an example method for
obtaining an ECR waveform.
[0026] FIG. 2A is a schematic diagram depicting operation of an
example system for obtaining an ECR and for fitting the cochlear
stimulation system for use by the user.
[0027] FIG. 2B is a schematic block diagram of an example system
that may be used to implement the system illustrated in FIG.
2A.
[0028] FIG. 2C is a schematic block diagram of another example
system that may be used to implement the system illustrated in FIG.
2A.
[0029] FIG. 2D is an example display that may be generated using an
example of the systems illustrated in FIGS. 2A-C for analyzing
operation of a cochlear stimulation system.
[0030] FIG. 3 is a flowchart depicting operation of an example
method for fitting a cochlear stimulation system in a user.
[0031] FIG. 4 is a flowchart depicting operation of an example of a
method for calibrating the cochlear stimulation system.
[0032] FIG. 5 is a flowchart depicting operation of an example of a
method for obtaining a performance evaluation and failure detection
analysis of the user's cochlear stimulation system.
[0033] FIG. 6A is a set of graphs showing groups of ECR waveforms
at selected frequencies as a function of sound intensity.
[0034] FIG. 6B is another set of graphs showing groups of ECR
waveforms at selected frequencies as a function of sound
intensity.
[0035] FIG. 7A is a graph of ECR amplitudes against sound intensity
levels for four frequencies.
[0036] FIG. 7B is a graph of ECR amplitudes against sound intensity
levels for four frequencies.
[0037] FIG. 8 is a graph showing the ECR time lag, particularly of
the negative peak B, versus external sound stimulus frequency.
[0038] FIG. 9 is a graph illustrating the use of ECR measurement to
detect an improperly implanted cochlear stimulation system in a
patient.
[0039] FIGS. 10A-10B are flowcharts depicting operation of a method
for analyzing a collection of EEG signal epochs to detect ECR
waveforms.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
I. Introduction
[0040] The following describes examples of systems and methods for
fitting a cochlear stimulation system to a patient. The examples
described below provide non-invasive objective techniques for
fitting the cochlear stimulation system that may be performed with
a sedated or sleeping patient, on adults or children. The examples
also perform fitting techniques that involve adjusting the dynamic
range of psychophysical levels in response to actual sound, and not
to direct electrical stimulation of the intracochlear electrodes.
The examples may also be implemented for use with any cochlear
stimulation system that permits adjustment of psychophysical levels
during fitting. Even cochlear stimulation systems that do not
require fitting may make use of example systems and methods
described below in performance evaluation, fault detection or
audiometric threshold evaluation.
[0041] The examples below are described in the context of cochlear
stimulation systems that are fitted by setting a T level (and if
capable, a C level as well) in terms of a dynamic range of current
stimulation levels. However, it is to be understood that example
systems and methods may be implemented to perform fitting by
setting any psychophysical parameter(s) according to sound levels.
The examples are also described in the context of cochlear
stimulation systems having an external, non-implanted component
that contains a sound processor, and components for programming, or
mapping, and for implementing a stimulation strategy. However, it
is to be understood that the sound processor may be anywhere, even
in the implanted component, and examples of the systems and methods
described below may be used to provide fitting of such fully
implanted cochlear stimulation systems.
[0042] The example systems and methods described may be used to
perform a variety of functions. Such functions include: [0043] 1.
Fitting--the setting of the dynamic range of current stimulation
levels for each intracochlear electrode that is appropriate for the
individual patient; [0044] 2. Calibrating--the setting of the
dynamic range of current stimulation levels for each intracochlear
electrode using a known T level for a given frequency value as an
initial setting; [0045] 3. Performance Evaluation/Fault
Detection--for assessing the operation of the cochlear stimulation
system, or even for detecting faults in the implant procedure;
[0046] 4. Audiometric Threshold Estimation--for assessing the
implanted patient's hearing while using the cochlear stimulation
system.
[0047] It is to be understood that the list of functions above is
not an exhaustive list of functions that may be performed. These
functions are just examples of the many functions available to
physicians, clinicians or other professionals having patients that
use a cochlear stimulation system and are able to analyze the
patient's Electrical Cochlear Response ("ECR").
II. the Electrical Cochlear Response
[0048] The Electrical Cochlear Response ("ECR") is a measure of
electrical activity generated by the residual cochlear tissue in
response to an electrical stimulation that results when the
cochlear stimulation system processes an external sound. The ECR is
measured in response to an external sound having a known sound
intensity and frequency. The ECR measurement is taken as a function
of actual sound being processed by the sound processor components
of the cochlear stimulation system being fitted. Because the ECR is
measured using signals generated by the user's residual cochlear
tissue and in response to actual sound being processed by the
components of the cochlear stimulation system, the information
obtained by using the ECR are tailored to the user's particular
needs.
[0049] FIG. 1A is a graph depicting an example ECR waveform 100.
The ECR waveform 100 is an electrical potential over a time period
due to electrical current passing through an intracochlear
electrode whenever a patient's cochlear stimulation system
processes a sound. The ECR waveform 100 is characterized by
parameters ("ECR parameters") determined from patterns and
measurements that change in accordance with changes in sound
intensity and frequency of the input sound. The time-variant, ECR
waveform 100 may be picked up by using several electrodes used to
pick up EEG activity ("EEG electrodes") strategically placed on the
patient's head to receive the strongest possible nervous system
responses to sound. During the detection of an ECR waveform 100, a
sound having a known intensity and frequency is generated and
processed by the cochlear stimulation system. As the sound is being
processed, the patient's EEG signals are picked up at the EEG
electrodes and stored. The EEG signals contain the ECR waveform 100
if the sound is being processed and therefore perceived by the
patient. As such, the ECR waveform 100 includes electric potential
contributions from auditory nerve fibers activity, cochlear nucleus
and residual cochlear tissue in the intracochlear electrode
vicinity.
[0050] The EEG signals contain other types of signals that may be
much stronger than the ECR waveform 100. For example, signals
arising from neuromuscular activity, or other types of nerve and/or
brain activity, all of which may have stronger signals than the ECR
waveform 100, may also be part of the EEG signals. The ECR waveform
100 may be "extracted" from the EEG signals by recording the EEG
signals as multiple time segments of EEG signals picked up while
the patient is subjected to a sound with a known and fixed
intensity and frequency. The multiple time segments are then
averaged to reduce the effect that electrical activity not
associated with the sound has on the EEG signals. This process is
described in more detail below with reference to FIG. 1B.
[0051] ECR waveforms 100 may be recorded for each frequency band to
which an intracochlear electrode is assigned, and thus obtain an
ECR for each electrode. The sound intensity is adjusted according
to the function being performed.
[0052] The ECR waveforms 100 include ECR activity peaks having
measurable properties that can be grouped as temporal (latency and
time course), spatial (morphology, amplitude and phase) and
frequency properties. The changes in these properties may be
measured or detected as the sound intensity level, frequency, or
current stimulation levels are varied.
[0053] An individual ECR waveform 100 may include a negative
potential peak B sometimes followed by a positive potential peak C,
and followed by a negative potential peak D. The waveform 100
levels out along a basal line to approximately a zero potential
value. Amplitude and time relationship of these peaks are labeled
on the ECR waveform 100 in FIG. 1A as Amp.sub.B, Amp.sub.C,
Amp.sub.D and t.sub.B, t.sub.C, t.sub.D respectively. The ECR may
be obtained when the sound processor in the cochlear stimulation
system senses and processes an input sound. The EEG signals are
then detected at the EEG electrodes and processed to obtain the ECR
waveform 100.
[0054] In FIG. 1A, the Y-axis or amplitude is measured in micro
volts (".mu.V") and the X-axis is the time window duration measured
in milliseconds ("ms"). The ECR waveform 100 is characterized by
the following: [0055] 1. Point A is the ECR waveform 100 starting
point; [0056] 2. Peak B is the minimum negative peak following
point A; [0057] 3. Peak C is the maximum positive peak following
peak B; [0058] 4. Peak D is the minimum negative peak following
peak C; [0059] 5. t.sub.A is the elapsed time from a starting point
of an analysis window of time that contains the ECR waveform 100 to
the ECR starting point A; [0060] 6. t.sub.B is the elapsed time
from the starting point of the analysis window to the negative peak
B; [0061] 7. t.sub.C is the elapsed time from the starting point of
the analysis window to the positive peak C; [0062] 8. t.sub.D is
the elapsed time from the starting point of the analysis window to
the negative peak D; [0063] 9. t.sub.va is the analysis window;
[0064] 10. Amp.sub.B is the peak B amplitude; [0065] 11. Amp.sub.C
is the peak C amplitude; and [0066] 12. Amp.sub.D is the peak D
amplitude.
[0067] The ECR waveform 100 characteristics listed above change in
relatively predictable ways as the intensity and frequency of the
input sound changes. These changes in the characteristics reflect
the change in residual cochlear tissue behavior that occurs when
the patient perceives the changing sound characteristics (intensity
and frequency). By identifying the intensity at which the ECR
waveform 100 forms, a clinician may identify the threshold level
("T level"). By identifying the intensity at which the ECR waveform
100 starts to become distorted, the clinician may identify the
comfort level (C level). The ECR waveform 100 provides an objective
method for determining the T and C levels.
[0068] FIG. 1B shows a set of signals illustrating operation of a
system for obtaining an ECR waveform. FIG. 1B shows a set of
signals illustrating operation of an example method for obtaining
an ECR. FIG. 1B includes a first signal diagram 102, a second
signal diagram 104, a third signal diagram 106, and a fourth signal
diagram 108. The first signal diagram 102 is the input sound
stimulus, Ea.sub.M,N, generated in M epochs for each of N
intracochlear electrodes. Each of the N electrodes corresponds to
one of N frequency bands processed by the bandpass filters
connected to each intracochlear electrode. Each intracochlear
electrode is identified by number (electrode no. 1, electrode no.
2, etc.), which may be referred to as a channel number, between 1
and N, and assigned a corresponding frequency band. The signals in
the first signal diagram 102 in FIG. 1B are: [0069] Ea.sub.1,1--a
sound signal generated in epoch M=1 at a selected intensity and at
a frequency=f.sub.C, the center frequency of the frequency band
processed by electrode no. 1. [0070] Ea.sub.1,2--a sound signal
generated in epoch M=1 at a selected intensity and at a
frequency=f.sub.C, the center frequency of the frequency band
processed by electrode no. 2. [0071] Ea.sub.2,1--a sound signal
generated in epoch M=2 at a selected intensity and at a
frequency=f.sub.C, the center frequency of the frequency band
processed by electrode no. 1. [0072] Ea.sub.3,1--a sound signal
generated in epoch M=3 at a selected intensity and at a
frequency=f.sub.C, the center frequency of the frequency band
processed by electrode no. 1.
[0073] The second signal diagram 104 shows a control signal 104a,
which establishes the starting point of an analysis window that
defines the time duration of an epoch. The control signal 104a in
FIG. 1B is a pulse of any suitable pulse width that signals the
start of: (1) a sound duration time, t.sub.d, which is the time
duration of the input sound; (2) an interval time, t.sub.i, which
is the interval time between two consecutive sound stimuli; and (3)
the analysis window, t.sub.va, which is the time duration of each
epoch.
[0074] The third signal 106 in FIG. 1B is an EEG signal 106a picked
up at the EEG electrodes. The EEG signal 106a is the signal that is
detected and recorded for measurement. As the EEG signal 106a is
recorded, it is stored in epochs. The fourth signal 108 shows the
epochs as segments, se.sub.M,N having time duration of t.sub.va
within the time interval t.sub.i. When the epochs are recorded, the
data is processed by averaging the epochs at a given electrode and
at a given intensity (IS).
[0075] During operation, the input sound signal shown in the first
signal diagram 102 is generated as a series of tones, or "pips,"
having the indicated characteristics. For example, a first pip,
Ea.sub.1,1, is generated for a duration t.sub.d. Random sound, or
no sound, is generated for a duration of t.sub.i-t.sub.d. The
second pip, Ea.sub.1,2, is generated for a duration of t.sub.d
followed by no sound for t.sub.i-t.sub.d. The pips may be generated
in any order, or randomly. In FIG. 1B, the first pip, Ea.sub.1,1,
is for the first epoch (M=1) corresponding to the data being
collected for the electrode number 1. The next pip is for the first
epoch corresponding to the data being collected for electrode
number 2. The next pip in FIG. 1B is for the second epoch
corresponding to the data being collected for electrode number 1.
The next pip, Ea.sub.3,1, is the third epoch corresponding to the
data collected for the electrode number 1.
[0076] The pips are generated in whatever order is selected until
the desired number of epochs, M, are collected for each electrode,
the data is analyzed to determine if any ECR waveforms resulted
from the input sound stimulus. Using the conventions established in
the description above with reference to FIG. 1B, a measured
response to the sound stimulus may be defined to the result of the
averaging of the epochs collected at a given frequency and at a
given intensity. Thus, a measured response for a given
intracochlear electrode, j, may be expressed as SE fc.sub.e(j) in
EQN. 1:
SE ? = ? ? ( ? , ? ) M ? ? = 1 , 2 , , N ? indicates text missing
or illegible when filed EQN . 1 ##EQU00001##
[0077] The measured response, SE fc.sub.e(j), is a waveform formed
by the average value of the signal levels in the EEG epochs at time
increments within the analysis window. The data collected for a
given intracochlear electrode, j, is analyzed, either visually, or
using pattern recognition software.
[0078] In general, a visual inspection of the measured response may
entail heuristically searching for ECR characteristics based on the
following guidelines: [0079] t.sub.A is typically less than about
10 ms. [0080] t.sub.B is typically about 10.+-.2 ms. [0081] t.sub.C
is typically about 15.+-.2 ms. [0082] t.sub.D is typically about
29.+-.2 ms.
[0083] It is noted that these values may be typical for a test
performed in sound field conditions with the speaker placed one
meter away from the implanted patient. There may be differences in
the values based on the individual patient, test conditions, and
other factors. The values above are provided as an example and do
not represent absolute parameters to which any results should
conform. In addition, the clinician may inspect the values of
amplitudes Amp.sub.B, Amp.sub.C, and/or the difference between the
two (Amp.sub.C-Amp.sub.B). The clinician may determine desired
minimum values of the amplitudes based, for example, on a model ECR
waveform created by using historical data, such as measured
responses that were deemed to be ECR waveforms for low threshold
sound levels. The clinician compares the values of Amp.sub.B,
Amp.sub.C, and/or Amp.sub.C-Amp.sub.B with expected minimum values
of each to determine whether the measured response is indeed an ECR
waveform.
[0084] The heuristics described here for a visual inspection may
also be implemented, for example, in a computer program designed to
analyze the measured responses and determine if the responses are
ECR waveforms.
[0085] When the analysis determines that an ECR waveform has been
detected, the frequency of the pips generated to stimulate the ECR
response is used to determine the intracochlear electrode involved
in generating the ECR. The signal diagrams 102-108 in FIG. 1B
illustrate a general method for obtaining an ECR. The information
conveyed by the ECR varies according to the function being
performed. For example, fitting involves setting the sound input to
a desired low level for the particular patient and setting the
frequency to select one of the intracochlear electrodes. The
current stimulation level is then increased from a very low value
until the ECR is obtained. The current stimulation level at which
an ECR waveform was detected is then set as the T level. The C
level may be determined experimentally by increasing the current
stimulation level until the ECR waveform begins to show distortion.
The C level may also be set to a percentage of T above the T level:
C level=T level+% age of T level. The frequency is then adjusted to
select another electrode and the current stimulation level is again
adjusted to a low level to determine a T level (and C level) for
the next electrode. The process continues until each electrode has
been fitted.
[0086] In other functions, such as calibration of a cochlear
stimulation system that has been operating, the sound intensity is
set to a low inaudible level to begin with. The frequency is set to
select an intracochlear electrode and the intensity is increased
until an ECR is detected. This measured minimum sound level is
compared to a desirable minimum sound level. If a meaningful
difference exists between the desirable and the measured minimum
sound levels, the current stimulation level is adjusted until an
ECR waveform is detected for the desirable minimum sound level.
[0087] FIG. 1C is a flowchart illustrating an example method for
obtaining an ECR waveform. The method illustrated in FIG. 1C may be
performed once a patient has had a cochlear stimulation system
implanted, and is connected to a selected set of EEG electrodes. In
addition, the method illustrated in FIG. 1C may be used to perform
virtually any function made possible by measuring the ECR. For
example, the method in FIG. 1C may be part of a fitting process to
set T and C levels in a newly implanted cochlear stimulation
system. In the case of a fitting, the patient may be fitted while
sedated or asleep when the cochlear stimulation system is turned on
for the very first time a few weeks after implantation surgery. The
method may also be part of a method for calibrating a cochlear
stimulation system that has been in use. The method may also be
part of a method for evaluating the performance or detecting fault
in the system. Suitable systems for carrying out the method
illustrated in FIG. 1C are described in more detail below with
reference to FIGS. 2A-2C.
[0088] Once the cochlear stimulation system has been implanted, the
system is initialized. Part of the initialization process may be to
set the system with an initial set of T and C levels. Once
initialized, a sound intensity level and frequency is selected as
an initial set of test characteristics, as shown in Step 110. The
sound with the selected frequency and intensity level is generated
to be received as noise-free as possible by the sound pickup on the
cochlear stimulation system as shown at Step 112. At Step 114, the
epochs of EEG signal are acquired and stored in memory in a
computer that may be connected to the EEG device to receive data.
The epochs are keyed or indexed or otherwise organized to
correspond to a given electrode (and therefore frequency band), and
at the selected intensity level. The data organization may depend
on the function being performed. If the patient is being fitted,
Step 114 may be performed such that epochs are acquired at the
selected frequency and desired low threshold intensity level an ECR
is detected. The frequency is varied to fit each intracochlear
electrode.
[0089] Step 116 is the storing step in which the epochs are stored
in memory as described. Once the desired number of epochs has been
collected, the data is analyzed at Step 118 to detect the ECR as
described above with reference to FIG. 1B. Step 120 performs peak
detection and measurement as part of a pattern recognition
algorithm that may be applied to the epoch data. At step 122, the T
level may be automatically set by software control. Step 120 may
also be performed by displaying the ECR waveforms as a function of
either intensity, frequency or both. The clinician may then decide
on the basis of the ECR waveforms, which indicates a T level and
which indicates a C level. Step 122 would then involve setting the
T and C levels manually (or using the assistance of a computer)
according to the specifications of the specific cochlear
stimulation system.
[0090] In general, for the process of fitting the system, the
intensity is set to a low threshold level, and the current
stimulation level is increased until the epoch data indicates that
it contains an ECR waveform. When an ECR waveform can be discerned,
the clinician may note the frequency to identify the electrode and
determine the current stimulation level being generated for the
desired low threshold sound intensity level. The determined current
stimulation level may then be set as the T level for the given
electrode. In some cases, the clinician may also elect to
specifically set the C level (comfort level) to set a maximum level
for loudness. The clinician may determine the C level by increasing
the intensity until the ECR waveform becomes distorted. The next
lower level of intensity that produced an un-distorted ECR waveform
may be selected for determining the C level.
Iii. Example Systems for Fitting a Cochlear Stimulation System
[0091] FIG. 2A is a schematic diagram depicting operation of an
example system 200 for obtaining an ECR and for fitting a cochlear
stimulation system for use by the user. The system 200 in FIG. 2A
is described in the context of fitting the user with a cochlear
stimulation system, which includes a sound pickup 202 (such as a
microphone, or other auditory signal input device), a sound
processor 206, an implanted component 208, a transmitter 209, a
signal-carrying lead 210 and an intracochlear electrode array 212.
The system 200 for obtaining ECR includes a plurality of EEG
electrodes (or, scalp electrodes), an EEG acquisition device 216 to
output EEG signals, an ECR waveform processor 218, and a user
interface 220.
[0092] FIG. 2A illustrates operation of the system 200 beginning
with the generation of a sound stimulus 204 to be received by the
sound pickup 202. The sound pickup communicates the electrical
signals representing the sound to the sound processor 206, which
processes the sound by de-multiplexing the electrical signals
according to the frequency bands defining the bandpass filters in
the sound processor 206, and by selecting the current stimulation
level appropriate for the intensity of the sound. The signal is
then communicated to the transmitter 209, which transmits the sound
signal to a receiver in the implanted part 208. The implanted
portion 208 includes electronics for multiplexing the signal to
couple signals to the appropriate intracochlear electrode according
to the frequencies of the multiplexed signal. The signals are
carried over the signal-carrying lead 210 to the intracochlear
electrode array 212.
[0093] As shown in FIG. 2A, the cochlear stimulation system
operates as intended by processing the sound generated at the sound
input 202. The characteristics of the sound may be controlled by,
for example, controlling the intensity of the sound as well as the
frequency of the sound. During operation of the cochlear
stimulation system, the EEG acquisition device 216 pickups up EEG
signals from the EEG electrodes 214. The EEG electrodes 214 include
four electrodes. In an example implementation, the four EEG
electrodes may include electrodes identified as A.sub.1, A.sub.2,
Cz, and FP.sub.z according to a known convention for identifying
EEG electrodes.
[0094] The EEG electrodes 214 may be placed on any part of the body
from which the strongest possible EEG signals may be picked up. In
general, the locations of the EEG electrodes 214 will be on the
patient's head. Two of these EEG electrodes, A.sub.1 and A.sub.2,
are relative references, one electrode is the active or positive
and the fourth electrode is the common or ground. The main EEG
signal electrodes and are typically placed near the right and left
ears.
[0095] As the sound input 204 is received at the sound pickup 202
and processed by the cochlear stimulation system, the EEG
acquisition system 216 records the EEG signals to obtain a picture
of what the residual tissue inside the cochlea looks like when it
is being stimulated by a electrical current whenever sound
processor processes a sound. The EEG signals are processed at the
ECR waveform processor 218 by averaging the EEG signals in epochs
as described above with reference to FIG. 1B. FIG. 2A shows the
right and left EEG signals 218a & b, a stimulus signal 218c,
and a control signal 218d. The resulting ECR waveform information
may be displayed on a user interface 220, which shows a display
possible in both the time and frequency domain.
[0096] FIG. 2B is a schematic block diagram of an example system
that may be used to implement the system illustrated in FIG. 2A.
The system 230 in FIG. 2B includes an audiometric enclosure 234 in
which the patient is fitted with a cochlear stimulation system 240,
an ECR detection system 260, and an input sound generator 280. The
patient in the audiometric enclosure 234 is fitted with a set of
EEG electrodes 232, which are used for picking up the implanted
patient's EEG activity. The'EEG electrodes 232 may include four EEG
or "scalp" electrodes 236, for example placed on the scalp of the
patient 246. The EEG electrodes 232 are connected via the scalp
electrodes 236 to the ECR detection system 260.
[0097] The cochlear stimulation system 240 includes a sound
processor 248 connected to a microphone worn by the patient as an
earpiece via a communication link 256, an external
receiver/transmitter 238, an internal receiver/transmitter 242, and
an intracochlear electrode array 244. The microphone receives sound
from a sound field 252 and converts the sound to electrical
signals. The electrical signals are communicated to the sound
processor 248 via the communication link 256. The sound processor
248 processes the electrical signals according to a selected
stimulation strategy and communicates the processed signals to the
external receiver/transmitter 238, which may be via a wireless
link. The external receiver/transmitter 238 communicates the
processed signals to the internal receiver/transmitter 242. The
internal receiver/transmitter 242 is implanted in the patient's
head in a location that would permit communication with the
external receiver/transmitter 238. The internal
receiver/transmitter 242 includes electronics for processing the
signal received from the sound processor 248. The intracochlear
electrode array 244 is implanted inside the patient's cochlea and
connected to receive electrical signals from the internal
receiver/transmitter 242. Whenever one of the intracochlear
electrodes is activated, an electrical current is delivered to the
patient's auditory nerve. A surface (scalp) electrical potential or
voltage generated by this electrical current is picked up by the
EEG electrodes 236.
[0098] The patient, who may be sedated or asleep, is positioned in
the audiometric enclosure 234. The patient, while wearing cochlear
stimulation system 240 is positioned near the front of an audio
speaker 254, with the speaker 254 facing the microphone 248 of the
cochlear stimulation system 240. The audiometric enclosure 234 is
configured to be noise-free, or at least as noise-free as possible,
in a sound field 252 between the speaker 254 and the microphone 248
(worn by the patient).
[0099] The ECR detection system 260 includes an EEG acquisition
system 262, an ECR waveform processor 264, and an ECR output
processor 266. The EEG acquisition system 262 receives EEG signals
from the EEG electrodes 232 and sends the EEG signals to the ECR
waveform processor 264. The ECR waveform processor 264 performs the
averaging of the EEG epochs as described above with reference to
FIGS. 1A-1C. The ECR waveform information may be processed by the
ECR output processor 266. The ECR output processor 266 may include
a user interface that provides printing and display resources to
provide a clinician with a graphical representation of the measured
responses, which may include ECR waveforms. The ECR output
processor 266 may also include a process for automatically
detecting the ECR waveforms from the measured responses and may
also determine the desired information from the ECR waveforms. For
example, the ECR output processor 266 may include software such as
pattern recognition software to analyze the measured responses and
determine which if any are ECR waveforms. The software may also
determine which intracochlear electrode corresponds to the detected
ECR waveforms, and determine T and C levels for the intracochlear
electrode. The software may also include calibration, performance
evaluation and fault detection methods, similar to the example
methods described below with reference to FIGS. 3-6. The ECR output
processor 266 may also include a link (not illustrated) to the
cochlear stimulation system 240 to download the T and C levels
directly in the cochlear stimulation system 240. Such a link may be
via a wired connection, or a wireless connection.
[0100] In an example implementation, the EEG acquisition system 262
includes two channels with differential inputs operating in AC mode
with a gain of 12,500 or more, a bandwidth of about 30 to 500 Hz.
An A/D converter is used in an example implementation. The A/D
converter may be 10 bits resolution and have two channels with a
sampling rate of 20 kHz. Also in an example implementation, the ECR
waveform processor 264 and ECR output processor 266 may be
implemented using a general-purpose computer, or some other
computerized device, that implements examples of methods described
herein. In one example, the ECR waveform processor 264 may effect
averaging of up to 300 epochs of EEG in intervals of up to 75
ms.
[0101] The sound stimulus for performing the fitting, calibrating,
performance evaluation, etc. may be provided by the input sound
generator 280. The input sound generator 280 includes a signal
generator 286, an attenuator 284, and an amplifier 282. The signal
generator 286 generates a signal at a selected frequency. The
signal generator 286 may be programmed to generate the signal in a
desired pattern, such as a random sequence of pips, or in sequences
ordered in a desired way according to frequency. The programmed
signal generator 286 may include a sound level input and
communicate with the attenuator 284 and amplifier 282 to generate
the sound at a proper dB.sub.HL setting.
[0102] The attenuator 284 and amplifier 282 operate to keep the
signal-to-noise ratio (SNR) as low as possible, and to provide a
sense of a substantially linear relation between the output sound
level (in dB.sub.HL) and the signal voltage output from the
amplifier 282. The input sound generator 280 allows the clinician
to control the input signal by setting a frequency and a sound
intensity (in dB.sub.HL).
[0103] In an example implementation, the input sound generator 280
is capable of generating up to 90 dB.sub.HL at one meter from the
speaker 254 at a frequency range of 500 to 8000 Hz with a THD of
less than 2%, a tolerance of 1% from the nominal frequency, and can
be adjusted in increments of one, half, or third octaves.
[0104] FIG. 2C is a schematic block diagram of another example
system that may be used to implement the system illustrated in FIG.
2A. The system 231 in FIG. 2C is similar to the system 230 in FIG.
2B including the same components. One difference between the system
200 in FIG. 2B and the system 231 in FIG. 2C includes a test sound
chamber 257 of reduced dimensions. Sound from the speaker 254 is
coupled to the reduced dimensions test sound chamber 257 by a
waveguide 259. The sound field 245 inside the test chamber 257 is
calibrated to meet the same quality requirements as in the external
sound field 252 in FIG. 2B. The advantage of the reduced dimensions
test chamber 257 is that it eliminates the need to perform the
testing in a special room adapted to provide the desired noise-less
environment. The reduced test chamber 257 provides the desired
noise-less environment in a substantially portable chamber.
[0105] Once the ECR waveform processor 264 generates the ECR
waveform data, it may be analyzed by a clinician either by viewing
the ECR waveforms as a function of desired variables on a display,
or by pattern recognition or image processing software that
determines whether an ECR waveform is detected, or is becoming
distorted. FIG. 2D is an example display that may be generated
using an example of the systems illustrated in FIGS. 2A-C for
analyzing operation of a cochlear stimulation system. The display
in FIG. 2D is a series of sets of waveforms at six selected
frequencies (200 Hz, 500 Hz, 1000 Hz, 1500 Hz, 2000 Hz, 3000 Hz).
The sets of waveforms are plotted as a function of sound level (in
dB.sub.HL) on the vertical axis, and as a function of time on the
horizontal axis. The display in FIG. 2D illustrates the progression
through which the ECR waveform changes as the sound intensity is
increased for each given frequency. As FIG. 2D shows, each
intracochlear electrode generates ECR waveforms illustrating
different behavior at the selected sound intensity levels. For
example, a lowest intensity response signal 203 at 500 Hz. is
beginning to display an ECR waveform as low as 25 dB.sub.HL. A
lowest intensity response signal 205 at 1000 Hz doesn't begin to
show an ECR waveform. The ECR waveform at 1000 Hz doesn't begin to
appear until about 40 dB.sub.HL.
[0106] At the high intensity levels, an ECR waveform is clearly
present at 50 dB.sub.HL for a higher intensity response signal 207
at 1500 Hz and it appears that the patient may be able to perceive
the 1500 Hz sound louder than 3000 Hz sound, according to the
response signal amplitudes 207 and 209. In some display outputs,
the current stimulation level may also be displayed for the
results.
IV. Example Methods for Fitting a Cochlear Stimulation System
[0107] FIG. 3 is a flowchart depicting operation of an example
method for fitting a cochlear stimulation system in a user. Fitting
is a procedure typically performed when a user has a cochlear
stimulation system implanted. The example method described with
reference to FIG. 3 is an objective method performed while the
patient is sedated, or asleep, that may be performed to fit any
cochlear stimulation system. The patient typically undergoes a
surgical procedure to have the cochlear stimulation system
implanted. The procedure involves inserting the intracochlear
electrode array into the cochlea and connecting the intracochlear
electrode array to the device that receives signals from the sound
processor. The cochlear stimulation system is typically left in the
patient for a few weeks before the fitting procedure in an
unpowered state.
[0108] The fitting procedure may be performed using a system
similar to the example systems for obtaining an ECR described above
with reference to FIGS. 2A-2C. Once the patient is prepared for the
fitting inside the audiometric enclosure 234 (FIG. 2B), or in any
location if the fitting is to be performed using the reduced
dimensions chamber 257 (FIG. 2C), the system for obtaining the ECR
is activated (Step 300).
[0109] At step 302, the fitting method is initialized, which may
involve performing any procedure necessary to enable the system to
perform the fitting. As shown in FIG. 3 at step 302, the
initialization at least entails identifying information about the
cochlear stimulation system. This may be performed, for example, by
downloading, or requesting by manual input via a user interface,
information from the cochlear stimulation system's map. An example
of the type of data that is used in a fitting procedure is listed
in Table 1. Table 1 lists the intracochlear electrodes e(j), j=1 .
. . N, N=Number of active electrodes, the frequency band assigned
to the electrode, the center frequency (fc), the T.sub.j level (low
threshold current stimulation level) for electrode e(j), the
C.sub.j level (high threshold current stimulation level) for
electrode e(j), the low threshold sound intensity IS.sub.MIN that
correspond to T level, and the high threshold sound intensity
IS.sub.MAX that correspond to C level. At step 302, these values
may be set to initial (e.g. factory settings) values. Once the
fitting procedure is complete, the T and C values will be set to
values particular to the patient. At step 302, a variable N is set
to the number of intracochlear electrodes and for each electrode,
the center frequencies, f.sub.ce(j), are retrieved.
TABLE-US-00001 TABLE 1 Intracochlear Frequency f.sub.ce(j) Tj Cj
IS.sub.MIN IS.sub.MAX electrode e(j) Band (Hz) (Hz) (CU) (CU)
(dB.sub.HL) (dB.sub.HL) 1 6938-7938 6988 135 199 25 75 2 6063-6938
6500 137 199 30 80 3 5313-6063 5688 138 200 25 80 . . . N = 22
188-313 250 133 194
[0110] At step 304, an initial current stimulation level is set to
A for each electrode. The value A should be a low current
stimulation level. Current stimulation values in typical cochlear
stimulation systems are between 0 and 255 Clinic Units (CU), where
the lower values reflect lower current levels. Units and ranging
values for measuring stimulation current may vary depending on
cochlear stimulation system manufacturer. In this example, the
value A should be a very low level since an objective of the
fitting process is to determine a current level that permits the
user to hear a desired low threshold sound level. Thus, the other
parameter that is initialized in step 304 is a desired low
threshold sound level, IS.sub.TH.
[0111] At step 306, a sound stimulation signal is configured for
use as the input sound signal during the fitting process. The sound
stimulation signal includes center frequencies corresponding to
intracochlear electrodes that are to be tested and a sound
intensity. For purposes of fitting, the sound intensity is kept
constant at the desired low threshold sound intensity level.
[0112] Initially, the sound stimulation signal includes a sequence
of pips at each of the center frequencies of the N intracochlear
electrodes. Each of the N center frequencies is generated M times.
The sound stimulation signal may be configured to emit the pips in
sequence as described above with reference to FIG. 1B. Each pip is
to be output for a time duration, t.sub.d. A pip is output in time
intervals of time t.sub.i, the start of which may be triggered by a
control signal that may be used as a sync to the EEG acquisition
system. In an example implementation, the center frequencies may be
stored in an array with an initial index of N. Once the sound
stimulation signal is configured, the fitting procedure continues
and eventually cycles back through step 306. Each time the fitting
procedure cycles through step 306, N is lower by one and a center
frequency has been removed from the next configuration of the sound
stimulation signal. At step 306, the remaining center frequencies
are re-grouped for the next sound stimulation signal.
[0113] At step 308, the configured sound stimulation signal is
output to enable the cochlear stimulation system to input the
sound. As the sound stimulation signal is being output, the
cochlear stimulation system is processing the sound and exciting
the intracochlear electrodes corresponding to the center
frequencies of each pip being generated. In addition, the patient's
EEG signals are detected and recorded to determine the nervous
system's response to the operation of the cochlear stimulation
system. The EEG signals are input and stored for processing. In an
example implementation, segments of EEG signal corresponding to the
emission of individual pips are stored in groups corresponding to
the frequency of the pip. The segments are stored in a memory
storage having the capacity to contain the signal levels sampled at
a selected sampling rate for a time equal to the analysis window,
t, (See FIG. 1B).
[0114] Each group of segments, which is referred to below as
se.sub.M,N, contains M segments such that there are M segments at
each of the N center frequencies.
[0115] At step 310, the EEG signals are analyzed to determine if an
ECR waveform may be detected for any of the center frequencies
generated in the sound stimulation signal. In this example, the
analysis involves averaging the segments at each given center
frequency to generate a measured response, SEfc.sub.e(j), using for
example, EQN. 1 according to techniques described above with
reference to FIGS. 1A and 1B. The measured responses at each
electrode, SEfc.sub.e0), are then process to measure the ECR
characteristics, such as the ECR characteristics described above
with reference to FIG. 1A. For each intracochlear electrode j
between 1 and N, at the given sound intensity (IS), the analysis
seeks to detect, measure, and store the ECR characteristics shown
in Table 2. Table 2 shows the ECR characteristics with expected
values for each. It is noted that the expected values shown in
Table 2 are examples of values that have been determined
empirically through experimentation. The values are provided for
purposes of illustration. Actual values may vary depending on the
patient, equipment used and knowledge gained from continued study
of ECR waveforms.
TABLE-US-00002 TABLE 2 ECR Expected Characteristic Value
Description t.sub.A <10 ms. Maximum latency of point A t.sub.B
10 .+-. 2 ms. Approximate time to Peak B t.sub.C 15 .+-. 2 ms.
Approximate time to Peak C t.sub.D 29 .+-. 2 ms. Approximate time
to Peak D Amp.sub.Bj PEAKB.sub.MIN Minimum absolute value of Peak B
amplitude Amp.sub.Cj PEAKC.sub.MIN Minimum absolute value of Peak C
amplitude Amp.sub.Dj PEAKD.sub.MIN Minimum absolute value of Peak D
amplitude |ECRj| = |ECR.sub.MinThresh| Minimum threshold for
absolute Amp.sub.Cj - Amp.sub.Bj value of the difference between
the Peak C amplitude and the Peak B amplitude
[0116] At step 310, the ECR characteristics for each measured
response, SEfc.sub.e(j), may be checked against an expected value,
such as the examples shown in Table 2. A heuristic based on the
expected values may be performed to determine whether a measured
response is an ECR waveform. Such a heuristic is represented in
FIG. 3 at decision block 312. The determination that a measured
response fits the pattern of the ECR waveform signifies that the
patient "hears" the tone at the center frequency, f.sub.ce(j), for
electrode e(j) at the desired minimum threshold sound intensity
level. It also signifies that the current stimulation level, A, has
reached a level sufficient to fire the auditory nerve to enable the
perception of hearing the tone. This may also be considered the
lowest current stimulation level for the electrode e(j), which is
therefore, the T level for electrode e(j).
[0117] At decision block 312, if it is determined that the measured
response is an ECR waveform, at step 318 the intracochlear
electrode associated with the frequency of the tone that generated
the response is identified. At step 320, the map of the cochlear
stimulation system is adjusted by setting the T level for e(j),
Tea), for electrode e(j) to the current stimulation level used to
generate the response, which is A. At step 322, the center
frequency of e(j), f.sub.ce(j), is removed from the set of
frequencies that are to be included in the next configuration of
the sound stimulation signal. At step 324, N is decremented to
indicate that one less intracochlear electrode is left to be
adjusted during the fitting. Decision block 326 checks to see if
the last intracochlear electrode has been adjusted. If N is not 0,
the next sound stimulation signal is configured at step 306. The
process then continues at steps 308, 310 and decision block 312 as
described above. If at decision block 326, N has reached 0, all of
the intracochlear electrodes have been adjusted by having the T
level set for each. At step 328, the C level for each electrode,
Ce(j), may be calculated as a function of the measured T level. For
example, the C level, Ce(j) may be determined as shown in step 328
using Ce(j)=Tea)+X %*Te(j). The C level is calculated for each
electrode e(j) where J=1, 2, 3, . . . , N. Once step 328 is
completed, a dynamic range of the current levels used for hearing
at the intracochlear electrodes has been determined.
[0118] The C level for the intracochlear electrodes may also be
determined by performing the fitting procedure again, except that
the initial current stimulation level, A, is initialized to be well
within an estimated dynamic range, which may be determined using
the equation for Ce(j) in step 328. The sound intensity level is
set to a desired maximum sound intensity. A procedure containing
steps similar to steps 306, 308, 310, 312, 318, 320, 322, 324, and
326 may be performed with modifications. For example, at the step
similar to step 310, the measured responses are processed to
determine if the ECR waveform is going away. That is, when the ECR
waveform is beginning to show distortion, the sound is becoming too
loud for the patient. At that point, the current stimulation level
A, or a value a few levels lower may be stored as the C level for
that electrode.
[0119] At step 330, intracochlear electrodes for which no response
was recorded are reported to the clinician. This may be via a
displayed message, or a printout, or by any suitable means for
generating error reports.
[0120] Referring back to decision block 312, if no ECR waveform is
detected among the measured responses, the current stimulation
level, A, is incremented by a predetermined incremental value A.
Step 314 calculates a new value of A as A=A+.DELTA.. At decision
block 316, the new current stimulation level, A, is checked against
a predetermined upper limit, A.sub.MAX. If the limit has been
exceeded, the process is halted and control proceeds to step 328.
If the upper limit has not been reached, the sound signal
stimulation that was used in the previous cycle that resulted in
the finding of no ECR waveform is re-used to see if the patient can
"hear" the signal, or any part of the signal, at a higher current
stimulation level.
[0121] Once the fitting process is completed successfully, the
patient may use the cochlear stimulation system to hear. It is
possible that after a period of continued use, the cochlear
stimulation system, or the patient, may change and result in a
change in the performance of the cochlear stimulation system. In
such case, the ECR waveform analysis may be used to calibrate the
cochlear stimulation system.
V. Example Methods for Calibrating a Cochlear Stimulation
System
[0122] FIG. 4 is a flowchart depicting operation of an example of a
method for calibrating the cochlear stimulation system. Examples of
methods for calibrating a cochlear stimulation system consistent
with those described here may be used to calibrate any cochlear
stimulation system.
[0123] In a calibration procedure, the patient is prepared in a
manner similar to the method for fitting described above with
reference to FIG. 3 using an example system for obtaining an ECR
such as those described above with reference to FIGS. 2B and 2C.
Once the patient is prepared for the fitting inside the audiometric
enclosure 234 (FIG. 2B), or in any location if the fitting is to be
performed using the reduced dimensions chamber 257 (FIG. 2C), the
system for obtaining the ECR is activated (Step 400).
[0124] At step 402, the fitting method may be initialized in a
manner similar to step 302 in the method for fitting in FIG. 3.
According to step 402, the initialization may entail identifying
the same information as in step 302 in FIG. 3. The example method
of calibration in FIG. 4 is described using conventions established
above in the description of the method of fitting. It is noted that
one difference between the example method of fitting and the
example method of calibrating is that the information retrieved in
step 402 in calibrating includes MAP information that may have
originated from a fitting procedure.
[0125] At step 404, the current stimulation level for each
electrode is set to A.sub.J, which is the current stimulation level
retrieved from the cochlear stimulation system MAP. In calibration,
the current settings of the MAP may be thought of as being
parameters under test to determine their effectiveness in creating
a perception of sound at a low sound intensity. At step 404, the
sound intensity level is set to a low initial setting. During
calibration, the sound intensity is increased in cycles while the
current stimulation levels remain constant. In general, the initial
sound intensity IS.sub.init should be set to a level at which one
is not expected to be able to hear.
[0126] At step 406, a sound stimulation signal is configured for
use as the input sound signal during the calibration process. The
sound stimulation signal includes center frequencies corresponding
to intracochlear electrodes that are to be tested and a sound
intensity. For purposes of calibration, the sound intensity is
initially as set above in step 404 and increased as the procedure
is performed. The configuration of the sound stimulation signal may
be performed in the same way as in step 306 in the method for
fitting. In addition, step 406 proceeds with one fewer center
frequency each time a cycle completes with the detection of an ECR
waveform similar to step 306.
[0127] At step 408, the configured sound stimulation signal is
output to enable the cochlear stimulation system to input the
sound. As the sound stimulation signal is being output, the
cochlear stimulation system is processing the sound and exciting
the intracochlear electrodes corresponding to the center
frequencies of each pip being generated. In addition, the patient's
EEG signals are detected and recorded to determine the nervous
system's response to the operation of the cochlear stimulation
system. The EEG signals are input and stored for processing. In an
example implementation, segments of EEG signal corresponding to the
emission of individual pips are stored in groups corresponding to
the frequency of the pip. The segments are stored in a memory
storage having the capacity to contain the signal levels sampled at
a selected sampling rate for a time equal to the analysis window,
t.sub.va (See FIG. 1B). Each group of segments, which is referred
to below as se.sub.M,N, contains M segments such that there are M
segments at each of the N center frequencies.
[0128] At step 410, the EEG signals are analyzed to determine if an
ECR waveform may be detected for any of the center frequencies
generated in the sound stimulation signal. The analysis may involve
averaging the segments at each given center frequency to generate a
measured response, SEfc.sub.e(j), as described for step 310 in the
method of fitting. At decision block 412, the measured responses
are processed using heuristics similar to those described above
with reference to FIG. 3 at step 310.
[0129] If at decision block 412, it is determined that the measured
response is an ECR waveform, the intracochlear electrode associated
with the frequency of the tone that generated the response is
identified at step 418. At step 420, the current sound intensity
level IS is identified as the minimum sound level perceived by the
patient at the intracochlear electrode e(j) determined to have
responded to the stimulation signal with an ECR waveform.
[0130] At decision block 422, the measured sound intensity is
compared with a sound intensity recognized as being a desirable
minimum level. If the measured minimum sound intensity IS.sub.J is
much greater than the desirable level, step 424 is performed. At
step 424, the electrode e(j) is reported as an electrode needing to
be re-fitted. Step 424 may involve identifying e(j) or marking it,
and then proceeding to step 426. The actual reporting may be
performed at the conclusion of the calibration method.
[0131] If at decision block 422, the difference between IS.sub.j
and a desirable minimum level is not significantly greater, the
center frequency, fc.sub.e(j), is eliminated from the center
frequencies that are to be used in the next configured sound
stimulation signal at step 426. At step 428, N is decremented to
indicate that one less intracochlear electrode is left to be tested
during the calibration. Decision block 430 checks to see if the
last intracochlear electrode has been tested. If N is not 0, the
next sound stimulation signal is configured at step 406. The
process then continues at steps 408, 410 and decision block 412 as
described above. If at decision block 430, N has reached 0, all of
the intracochlear electrodes have been tested at low threshold
levels.
[0132] At step 432, the C level for each electrode, Ce(j), may be
calculated as a function of the measured T level. For example, the
C level, Ce(j) may be determined as shown in step 432 using
Ce(j)=Te(j)+X %*Tea). The C level is calculated for each electrode
e(j) where j=1, 2, 3, . . . , N. Once step 432 is completed, a
dynamic range of the current levels used for hearing at the
intracochlear electrodes has been determined. At step 434, the
electrodes e(j) may be tested to determine if the C levels can
create a perception of hearing at a sufficiently high level. The
test for calibrating the cochlear stimulation system at a high
sound level may be performed in a manner similar to steps 406, 408,
410, 412, 418, 420, 422, 424, 426, 428, and 430 with modifications.
For example, at the step similar to step 410, the measured
responses are processed to determine if the ECR waveform is going
away. That is, when the ECR waveform is beginning to show
distortion, the sound is becoming too loud for the patient. At that
point, the sound level is compared with a desirable maximum sound
level. If the sound level is significantly less than the desirable
maximum, the electrode e(j) is designated as requiring
re-fitting.
[0133] Referring back to decision block 412, if no ECR waveform is
detected among the measured responses, the sound intensity level,
IS, is incremented by a predetermined incremental value .DELTA..
Step 414 calculates a new value of the sound intensity IS as
IS=IS+.DELTA.. At decision block 416, the new sound intensity
level, IS, is checked against a predetermined upper limit,
IS.sub.MAX. If the limit has been exceeded, the process is halted
and control proceeds to step 432. If the upper limit has not been
reached, the sound signal stimulation that was used in the previous
cycle that resulted in the finding of no ECR waveform is re-used at
the higher sound intensity, IS, to see if the patient can "hear"
the signal, or any part of the signal.
[0134] When the process for calibration ends, the electrodes that
require re-fitting and any that did not register an ECR are
reported. This may be via a displayed message, or a printout, or by
any suitable means for generating error reports.
VI. Example Methods for Using ECR for Evaluating Performance and
Detecting Failure
[0135] FIG. 5 is a flowchart depicting operation of an example of a
method for obtaining a performance evaluation and failure detection
analysis of the user's cochlear stimulation system. The example
method illustrated in FIG. 5 provides data relating to the response
of a cochlear stimulation system to sound inputs at low sound
levels. The data is captured, then analyzed for indications of the
performance of the cochlear stimulation system.
[0136] The example method illustrated in FIG. 5 is similar to the
method shown in FIG. 4 for calibrating the cochlear stimulation
system. The results of stimulating the cochlear stimulation system
to produce ECR waves at a low sound intensity may be analyzed for
anomalies that are indicative of a fault, or some other problem
with the system.
[0137] The patient may be prepared for the method illustrated in
FIG. 5 in a manner similar to the example methods of fitting and
calibration described above. The system is activated at step 500.
Steps 502, 504, 506, 508, 510, 512, 514, and 516 may be performed
in the same manner as steps 402, 404, 406, 408, 410, 412, 414, and
416 in FIG. 4. In step 504, the initial sound intensity is set to a
low level, L dB.sub.HL. Otherwise, the description of the operation
of steps 502, 504, 506, 508, 510, 512, 514, and 516 are described
above with reference to FIG. 4.
[0138] At step 518, the measured responses are analyzed to detect
all of the intracochlear electrodes for which an ECR was
registered. All m of the N electrodes for which an ECR waveform was
detected are identified. At step 520, the frequencies of the
detected intracochlear electrodes are removed from the next sound
stimulation signal. Decision block 522 determines if there are
anymore electrodes that have not registered a response. Once all of
the intracochlear electrodes have registered an ECR, or the sound
intensity has exceeded a maximum, the data may be displayed in a
variety of ways to determine if the cochlear stimulation system is
performing properly. Examples of the types of information that may
be displayed from the ECR waveform analysis include: [0139] ECR
amplitude (|ECR|) v. Sound pressure levels in dB.sub.HL [0140] ECR
time latencies v. frequencies of sound stimulation [0141]
Sufficiency of current stimulation levels in the dynamic range
[0142] Audiometric Threshold Estimation [0143] Identification of
intracochlear electrodes having ECR>ECR.sub.MAX [0144]
Identification of intracochlear electrodes that registered no ECR.
[0145] Identification of proper insertion of the intracochlear
electrodes into the cochlea. [0146] Identification of intracochlear
electrodes being over-stimulated.
[0147] The data collected during the analysis, the measured
responses to sound stimulation signals and detection of ECR
waveforms may be displayed in different ways to determine the
desired information. For example, the display shown in FIG. 2D
shows a series of measured responses at each of frequencies f=200
Hz, 500 Hz, 1000 Hz, 1500 Hz, 2000 Hz, and 3000 Hz. The series of
measured responses are for signals at different sound intensities
from lowest to highest. The audiometric threshold at each frequency
is the lowest sound intensity that generates an ECR waveform.
[0148] The method illustrated in FIG. 5 focuses on obtaining
measured responses at low levels. In another implementation, the
measured responses may be collected for a larger range of levels.
For example, the measured responses may be collected for sound
intensity levels through the dynamic range of sound levels
corresponding to the dynamic range of the current stimulation
levels of the cochlear stimulation system. The resulting measured
responses may be analyzed in a variety of ways to determine how the
cochlear stimulation system behaves through its dynamic range. FIG.
6A shows four series of measured response plots at frequencies
f=500 Hz, 1000 Hz, 2000 Hz, and 3000 Hz. Each series of measured
responses were generated by different sound intensities so that
several ECR waveforms may be analyzed. The ECR waveforms may be
analyzed to determine how the ECR amplitude, |ECR.sub.j|, changes
as the sound intensity increases. The change in time latency
between Peak B and Peak C as sound intensity increases is also
indicative of the cochlear stimulation system behavior. In general,
the ECR amplitude, |ECR.sub.j| should increase gradually as the
sound intensity increases. The time latency between Peak B and Peak
C should decrease gradually as well. Too sharp an increase in ECR
amplitude and too sharp a decrease in Peak B to Peak C latency may
indicate overstimulation by the intracochlear electrodes, which may
be corrected by calibrating the cochlear stimulation system. Too
gradual an increase in ECR amplitude and to gradual a decrease in
Peak B to Peak C latency may indicate understimulation.
[0149] The results shown in FIG. 6A have been determined to reflect
a properly stimulated cochlear stimulation system. The results in
FIG. 6A may be contrasted with those of FIG. 6B, which provides an
example of an over-stimulated cochlear stimulation system.
[0150] FIGS. 6A & 6B graphically display results from analyzing
measured responses and displaying the results as shown. The
analysis of the results leading to the conclusions indicated by the
displays in FIGS. 6A & 6B may also be processed by a computer
program that performs the analysis and outputs the conclusion
regarding whether or not the intracochlear electrodes are
over-stimulated or properly stimulated. For example, a method for
detecting ECR similar to the ECR detector method described below
with reference to FIG. 10A may be used to analyze measured
responses in the cochlear stimulation system's dynamic range of
sound intensity. A pattern recognition program, or a method similar
to the ECR peak analyzer described below with reference to FIG. 10B
may be used to measure peak characteristics at various sound
intensities and frequencies to determine whether the cochlear
stimulation system is over-stimulated.
[0151] The results may be viewed in other ways. FIG. 7A is a graph
of ECR amplitude growth function for four intracochlear electrodes
of the implanted patient's cochlear stimulation system. The graph
shows ECR amplitude, which is the Peak C and Peak B difference in
amplitude, versus external intensity sound stimulation for four
different frequencies f=500 Hz, 1000, Hz, 2000 Hz, and 3000 Hz. The
Y-axis is the amplitude measured in micro volts (".mu.V"), and the
X-axis is the sound levels measured in decibels ("dB.sub.HL"). A
linear regression may be performed on the amplitude showing a
linear plot for frequency. The slope (m) of each line indicates the
progression of the increase in ECR amplitudes with the increase of
sound intensity levels. The slopes are also indicated in FIG. 7A
for each frequency.
[0152] FIG. 7B shows a graph that is similar to that of FIG. 7A.
Where the graph in FIG. 7A provides results for a properly
stimulated cochlear stimulation system, the graph in FIG. 7B
provides results for an overly stimulated cochlear stimulation
system. The over-stimulation is evidenced by the higher slopes (m)
of the lines in FIG. 7B.
[0153] It is noted that while the results shown in FIGS. 7A &
7B are clear in the graphical display of the data, similar results
may be obtained automatically using a computer program to obtain
ECR peak characteristics and compare the ECR peak characteristics
to expected values. For example, data collected over time
consisting of graphs such as the graph in FIG. 7A may be
statistically analyzed to determine expected values for ECR
characteristics and for the slopes of the lines plotted in FIGS. 7A
& 7B. An ECR detector such to the ECR detector described below
with reference to FIG. 10A may be used to extract measured
responses indicating an ECR waveform in a collection of data taken
over the dynamic range of the sound intensity for the cochlear
stimulation system. An ECR waveform analyzer similar to the ECR
waveform analyzer described below with reference to FIG. 10B may be
used to obtain the ECR peak characteristics, and a test program may
be used to analyze conclusions about the peak characteristics, and
the slope of the lines plotted in FIG. 7B may be compared to
expected slope values to determine whether the cochlear stimulation
system is properly stimulated. Conclusions may be reached on the
same basis about whether a system is being under-stimulated as
well.
[0154] FIG. 8 provides another display of ECR waveforms that shows
how the time latency to Peak B (t.sub.B) changes with an increase
in frequency. FIG. 8 depicts a series of ECR waveforms of an
implanted patient for external sound stimulation of 50 dB HL at
frequencies from 250 to 6,000 Hz. As shown in FIG. 8, the time
latency t.sub.B increases as the frequency of the input signal is
decreased. The graph in FIG. 8 reflects normal operation. The
cochlea forms a spiral beginning at the oval window and ending at
an apex at the conical tip of the modiolus. High frequency signals
are sensed by the tissue closest to the oval window. Lower
frequency sounds are processed by tissue extending progressively
away from the oval window until the apex where the lowest
frequencies are processed. The graph in FIG. 8 therefore reflects a
properly implanted intracochlear electrode array.
[0155] FIG. 9 reflects another series of ECR waveforms obtained and
plotted in the same manner as the graph of FIG. 8. FIG. 9 does not
reflect a time latency, t.sub.B, that increases as the frequency
decreases. The graph in FIG. 9 reflects an improperly installed
intracochlear electrode array.
[0156] It is noted that while the results displayed in the graphs
in FIGS. 8 and 9 are output to a display for analysis. A computer
program may also be used to process the data from measured
responses to arrive at the same conclusions. An ECR waveform
analyzer similar to the method described below with reference to
FIG. 10B may be used to process the measured responses through the
entire sound dynamic range of the cochlear stimulation system. The
time latency at t.sub.B may be compared across the frequency range
used to obtain the measured responses and insure that the time
latency increases as the frequency is lower, which is as
expected.
VII. Automated ECR Waveform Analysis
[0157] FIGS. 10A-10B are flowcharts depicting operation of a method
for analyzing a collection of EEG signal epochs to detect ECR
waveforms. The methods illustrated in the flowcharts in FIGS. 10A
and 10B provide automated ECR waveform analysis and may be
configured to automate processes of fitting, calibration,
performance evaluation, fault detection, or any other procedure
relating to operation of the cochlear stimulation system. The
methods may be implemented in any computer-controlled device
configured to receive data obtained from measuring a patient's
electrical responses to operation of an implanted cochlear
stimulation system in a sound field created by a sound stimulus
signal. The examples illustrated by FIGS. 10A and 10B are described
in the context of data collected using examples of systems
described above with reference to FIGS. 2A-2C. However, the data
may be obtained using any system or method for any cochlear
stimulation system. The systems described above with reference to
FIGS. 2A-2C may be provided with hardware and software as an ECR
detector to perform examples of methods described in FIG. 10A, and
as an ECR peak analyzer to perform methods described in FIG.
10B.
[0158] Referring to FIG. 10A, a method for detecting ECR waveforms
is initiated at step 1000 when a desired collection of data is
available. The collection of data includes data sufficient for a
desired objective. For example, in a fitting, the collection of
data may include data representing responses at all frequencies of
operation of the cochlear stimulation system and all intensities
within a maximum desired dynamic range of sound intensities. In a
performance evaluation, the collection of data may be more limited.
For example, in audiometric threshold estimation, only a lower
range of sound intensities may be needed.
[0159] The system used to perform the example methods illustrated
in FIGS. 10A and 10B may be integrated with a computer-controlled
device designed to carry out any combination of the functions
performed by the components in the ECR detection system 260 in
FIGS. 2B and 2C in some example implementations. In such cases,
step 1002 may be performed as an integrated data input step in an
automated system for fitting and analyzing cochlear stimulation
systems. In other examples, the methods illustrated in FIGS. 10A
and 10B may be performed on a general-purpose computing platform,
which may receive the collected data by download, or by reading
from a data-storage medium, such as a CD, tape, portable storage
drive, or any other suitable medium.
[0160] At step 1002, the frequencies and intensities used in the
sound stimulus signal generated in collecting the data are input or
identified. The frequencies and intensities used may be obtained
from the response data collected. At step 1004, the frequencies and
intensities input in step 1002 are counted to determine the total
number of each as I.sub.max=total number of intensity values input,
and F.sub.max=total number of frequency values. At step 1006, the
frequency values and intensity values are ordered from minimum to
maximum. At step 1008, indices are initialized for the collections
of frequency and intensity values. An index f is initialized to f=1
to address the first frequency value in the collection of
frequencies. An index is initialized to i=1 to address the first
intensity value in the collection of intensities.
[0161] At step 1010, a collection of m epochs at frequency F(f) and
intensity I(i) are input from the response data. At step 1012, the
epochs are organized in k groups of n epochs such that k=m/n. A
partial average P.sub.A is calculated by averaging the epochs in
each group k of the m epochs: P.sub.A=Average of n epochs in group
A, for each A=1, . . . m/n. The values of m, n, and k are
preferably even numbers.
[0162] At step 1014, a parallel analysis may be performed on the
same set of data. In an example implementation, a known statistical
processing method such as, for example, the Fisher Single Point
(FSP) method, may be used to analyze the collection of m epochs at
frequency F(f) and intensity I(i). The selected method may set a
variable D.sub.E to indicate either true (T) or false (F) in
relation to whether or not an ECR waveform is detected.
[0163] At step 1016, a correlation, R.sub.xy(P.sub.A,P.sub.A+1), is
calculated of the partial average of each consecutive pair of
groups of k epochs, P.sub.A and P.sub.A+.sub.1. At step 1018, the
resulting set of correlation values is analyzed to determine a mean
of the correlation values greater than a selected upper threshold
correlation value for the k partial averages, P.sub.A.
[0164] At step 1020, a probability, P.sub.ECR, that the m epochs
are an ECR waveform is calculated based on the correlation values,
R.sub.xy and on the variable D.sub.E. At decision block 1022, the
probability, P.sub.ECR, is compared to a threshold probability. If
probability, P.sub.ECR, is greater than or equal to the threshold
probability, an ECR waveform is detected in the epochs collected
for frequency f and intensity i. The process continues at step 1024
for identification of the intensity as the minimum sound intensity
for which an ECR waveform is detected at frequency f. The frequency
f and intensity i are stored with an indication of detection of ECR
for the corresponding EEG epochs. At step 1026, the next frequency
is selected by setting index f=f+1. At decision block 1028, the
index f is checked to determine if the last frequency in the group
of frequencies has been analyzed. If f is not greater than
F.sub.Max, the index i is reset back to the first intensity (i=1)
of the group of intensities at step 1032. The process then
continues at step 1010. If the last frequency has been analyzed,
the results are reported for each frequency at which an ECR was
detected at step 1030.
[0165] Referring back to decision block 1022, if the probability of
ECR, P.sub.ECR, is less than the threshold, the non-detection of an
ECR waveform is indicated at step 1034. At step 1036, the intensity
index i is incremented to analyze epochs collected at the next
intensity. At decision block 1038, the index i is checked to
determine if the last of the intensities (I(i)>I.sub.MAX) has
been reached. If I(i)>I.sub.MAX, step 1040 is processed to
indicate that no ECR waveform was detected for the frequency f, and
step 1026 is performed to continue processing at the next
frequency. If at decision block 1038, I(i).ltoreq.I.sub.MAX, the m
epochs for the new intensity value at i and frequency f are
analyzed for an ECR waveform at step 1010.
[0166] The example method illustrated in FIG. 10A may be performed
to detect ECR waveforms at low thresholds of intensity, and may be
sufficient for performing fitting, calibration, audiometric
threshold estimation, or other functions in which the low threshold
of intensity at selected frequencies, is of interest. The results
may be reported on a display, printer, or stored for use by another
function, such as for example, a function for automatically setting
the dynamic range of the cochlear stimulation system based on the
ECR detection.
[0167] FIG. 10B illustrates an example of a method for analyzing
ECR peaks and latencies to provide information regarding the
performance or to detect faults in the cochlear stimulation system.
The example method in FIG. 10B is initiated at step 1042 after
performing an example of the method in FIG. 10A. The method in FIG.
10B processes data collected for a wider range of sound
intensities. At step 1044, the data corresponding to the
frequencies and minimum intensities is collected. In addition, for
each frequency, data corresponding to intensities greater than the
minimum intensity is also collected. At step 1046, a frequency
index, f, is initialized to f=1 to input the averages of the EEG
epochs where the first element is the frequency at f=1 at which an
ECR waveform was first detected. At step 1048, a collection is made
of the averages of the EEG epochs corresponding to the minimum
sound intensity at which an ECR waveform is detected, or greater,
for the frequency f. At step 1050, the averages of EEG epochs at
frequency f are organized by descending intensity.
[0168] At step 1052, the number of frequencies at which an ECR
waveform is detected is counted. At step 1054, a data enhancement
technique is performed on the ECR waveform to improve the SNR. Such
a data enhancement technique may include curve fitting or smoothing
techniques, or other techniques, that may interpolate the data. At
step 1056, the maximum and minimum values within each measured
response (averaged epochs within a time segment, e.g. analysis
window) at each sound intensity value are determined and stored. At
step 1058, the times t.sub.min and t.sub.max are determined. The
time t.sub.min is the time latency of Peak B, which is the time at
which the minimum amplitude, Peak B, is reached. The time t.sub.max
is the time latency of Peak C, which is the time at which the
maximum amplitude, Peak C, is reached. At step 1060, the ECR
characteristic parameters t.sub.B, t.sub.C, AmpB and AmpC are set
as follows: [0169] t.sub.B=t.sub.mm [0170] t.sub.C=t.sub.max [0171]
AmpB=min Amplitude [0172] AmpC=max Amplitude
[0173] At step 1062, the ECR characteristic parameters, t.sub.a,
t.sub.C, AmpB and AmpC, are stored as characteristics for the ECR
waveform detected at the given frequency f and for each intensity
i. At step 1064, the data for the next frequency is analyzed by
setting the index f=f+1. Decision block 1066 checks the index f to
determine if it is greater than the total number of frequencies, F.
If it is, then processing ends. If it is not, the next frequency is
analyzed, starting at step 1048.
[0174] While the present invention has been described with
reference to certain embodiments, it will be understood by those
skilled in the art that various changes can be made without
departing from the scope of the present invention. It will be
understood that the foregoing description of an implementation has
been presented for purposes of illustration and description. It is
not exhaustive and does not limit the claimed inventions to the
precise form disclosed. Modifications and variations are possible
in light of the above description or may be acquired from
practicing the invention. The claims and their equivalents define
the scope of the invention.
* * * * *