U.S. patent application number 11/142767 was filed with the patent office on 2005-10-06 for method of reading out the sensor elements of a sensor, and a sensor.
Invention is credited to Busse, Falko, Conrads, Norbert, Rutten, Walter.
Application Number | 20050218332 11/142767 |
Document ID | / |
Family ID | 7924302 |
Filed Date | 2005-10-06 |
United States Patent
Application |
20050218332 |
Kind Code |
A1 |
Rutten, Walter ; et
al. |
October 6, 2005 |
Method of reading out the sensor elements of a sensor, and a
sensor
Abstract
The invention relates to a method of reading out the sensor
elements of a sensor (1) with a matrix of light-sensitive or
X-ray-sensitive sensor elements (S.sub.1,2; S.sub.1,2 . . . ) which
are arranged in rows and columns and generate charges in dependence
on the incident quantity of radiation, the switches (3) of the
relevant sensor elements being activated via address lines (4, . .
. , 8, . . . ) and the charges of the respective activated sensor
elements being drained via read-out lines (9, 10, 11, . . . ) so as
to be processed further by way of amplifiers (14, . . . , 18, . . .
) and transfer means (19). The invention also relates to a
corresponding sensor as well as to an X-ray examination apparatus
which includes an X-ray source for emitting an X-ray beam for
irradiating an object in order to form an X-ray image, as well as a
detector for generating an electrical image signal from said X-ray
image. Despite the small incident X-ray doses, adequate amounts of
charge, and hence electrical signals, are provided so as to form
the image in that ingoing address lines (4, . . . , 8, . . . ) are
selectably connected, by way of individually controllable switch
elements (27, . . . , 30, . . . ) and by means of a switching
operation, to each time the respective next address line and the
sensor elements of at least two neighboring lines are activated by
means of one ingoing signal, and corresponding outgoing read-out
lines (9, . . . , 13, . . . ) are selectably connected, by way of
individually controllable switch elements (31, . . . , 34, . . . )
and by means of a switching operation, to the respective next
read-out line in such a manner that the charge signals read out
from the activated sensor elements of at least two neighboring
columns are combined so as to form one output signal.
Inventors: |
Rutten, Walter; (Linnich,
DE) ; Busse, Falko; (Aachen, DE) ; Conrads,
Norbert; (Raeren, BE) |
Correspondence
Address: |
PHILIPS INTELLECTUAL PROPERTY & STANDARDS
P.O. BOX 3001
BRIARCLIFF MANOR
NY
10510
US
|
Family ID: |
7924302 |
Appl. No.: |
11/142767 |
Filed: |
June 1, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11142767 |
Jun 1, 2005 |
|
|
|
09670585 |
Sep 28, 2000 |
|
|
|
Current U.S.
Class: |
250/370.09 ;
348/E3.02; 348/E5.086 |
Current CPC
Class: |
H01L 2924/0002 20130101;
H01L 2924/00 20130101; H04N 5/374 20130101; H01L 2924/0002
20130101; H04N 5/32 20130101; H04N 5/347 20130101 |
Class at
Publication: |
250/370.09 |
International
Class: |
G01J 001/00; G01T
001/24; H01L 025/00; H01L 027/00 |
Foreign Application Data
Date |
Code |
Application Number |
Oct 2, 1999 |
DE |
19947536.9 |
Claims
1. A method of reading out the sensor elements of a sensor (1) with
a matrix of light-sensitive or X-ray sensitive sensor elements
(S.sub.1,2; S.sub.1,2 . . . ), which are arranged in rows and
columns and generate charges in dependence on the incident quantity
of radiation, the switches (3) of the relevant sensor elements
being activated via address lines (4, 8, . . . ) and the charges of
the respective activated sensor elements being drained via read-out
lines (9, 10, 11, . . . ) so as to be processed further by way of
amplifiers (14, . . . , 18, . . . ) and transfer means (19),
characterized in that ingoing address lines (4, . . . , 8, . . . )
are selectably connected, by means of individually controllable
switch elements (27, . . . , 30, . . . ) and by way of a switching
operation, to the respective next address line in such a manner
that the sensor elements of at least two neighboring lines are
activated by means of one ingoing signal, and that corresponding
outgoing read-out lines (9, . . . , 13, . . . ) are selectably
connected to the respective next read-out line by means of
individually controllable switch elements (31, . . . , 34, . . . )
and by way of a switching operation, in such a manner that the
charge signals read out from the activated sensor elements of at
least two neighboring columns are combined so as to form one output
signal.
2. A method as claimed in claim 1, characterized in that the
magnitude of a group of sensor elements to be read out and/or the
distribution of a plurality of groups of sensor elements to be read
out across the overall matrix is predetermined as a binning pattern
and that the binning pattern is locally variable during an
exposure.
3. A method as claimed in claim 2, characterized in that it is
possible to intervene in the programmed execution of a binning
pattern during the exposure.
4. A sensor with a matrix of light-sensitive or X-ray sensitive
sensor elements (S.sub.1,1; S.sub.1,2 . . . ) which are arranged in
rows and columns and generate charges in dependence on the incident
quantity of radiation, with a respective switch (3) with an address
line (4, . . . , 8, . . . ) for each sensor line for connection to
activation means (20) via which the electrical sensor switches can
be activated, with a read-out line (9, . . . , 13, . . . ) for each
sensor column via which the charges of the respective activated
sensor elements are drained, with transmission means (19) at the
end of the relevant read-out line for converting the signals read
out in parallel into a serial signal, as well as with amplifiers
(14, . . . , 18, . . . ) which precede the transmission means,
characterized in that a first unit of switch elements (27, . . . ,
30, . . . ) is arranged between the activation means (20) and the
electrical sensor switches, that a second unit of switch elements
(31, . . . , 34, . . . ) is arranged between the electrical sensor
switches and amplifiers (14, . . . , 18, . . . ), and that there is
provided a system (21, 35) for controlling the switching operations
of the switch elements in order to connect each time a switch
element or a read-out line to at least one neighboring line.
5. A sensor as claimed in claim 4, characterized in that the switch
elements are arranged either on switch or read-out lines or are
integrated directly in the sensor matrix.
6. (canceled)
7. (canceled)
8. (canceled)
9. (canceled)
10. An X-ray examination apparatus, including an X-ray source (44)
for emitting an X-ray beam for irradiating an object so as to form
an X-ray image, and a detector (42) for generating an electrical
image signal from said X-ray image, characterized in that the X-ray
detector (42) includes a sensor as claimed in claim.
11. An apparatus for imaging radiation, comprising: a matrix
comprising: a plurality of sensor elements for sensing radiation,
wherein the plurality of sensor elements are configured in rows and
columns of the matrix, wherein each of the plurality of sensor
elements is adapted to generate a charge based on and in response
to sensing a quantity of radiation, wherein each row of sensor
elements comprises an electrical sensor switch operably coupled to
an activation means via an address line, whereby the activation
means is configured to activate the electrical sensor switch; a
plurality of readout lines, each operably coupled to at least one
column of the plurality of sensor elements and configured to drain
the at least one column's sensor elements which have been charged,
wherein each readout line comprises: a plurality of amplifiers,
each operably coupled to a readout line and adapted to amplify each
readout line's signal corresponding to at least one of the drained
charges; a transmission means operably coupled to the plurality of
readout lines for converting the drained charges into a serial
signal; a first unit having a plurality of switch elements,
operably coupled to the activation means and operably coupled to
the electrical sensor switches; a second unit having a plurality of
switch elements, operably coupled to the electrical sensor switches
and operably coupled to the amplifiers; wherein the switch elements
are configured to operate in coordination so that an output signal
includes combination of at least two neighboring readout line
signals.
Description
[0001] The invention relates to a method of reading out the sensor
elements of a sensor with a matrix of light-sensitive or X-ray
sensitive sensor elements which are arranged in rows and columns
and generate charges in dependence on the incident quantity of
radiation, the switches of the relevant sensor elements being
activated via address lines and the charges of the respective
activated sensor elements being drained via read-out lines so as to
be processed further by way of amplifiers and transfer means. The
invention also relates to a corresponding sensor as well as to an
X-ray examination apparatus which includes an X-ray source for
emitting an X-ray beam for irradiating an object so as to form an
X-ray image, as well as a detector for generating an electrical
image signal from said X-ray image.
[0002] Large-area X-ray detectors, consisting of a sensor matrix
with a plurality of sensor elements or pixels which are arranged in
rows and columns, are used in the field of medical X-ray diagnosis.
The sensor elements include photodiodes of amorphous silicon and
associated scintillator elements or, alternatively, photoconductors
for converting the X-rays into an electrical charge. Each sensor
element is provided with a switch for the reading out of the
charge, for example, with TFT (thin-film transistor) switches of
amorphous silicon or diode switches. The switches are closed or
activated so as to read out the collected charges conducted to
appropriate read-out lines wherefrom the charges flow into
appropriate charge-sensitive amplifiers which subsequently apply an
electronic signal to a multiplexer which applies the charge signals
from the relevant read-out lines to a data acquisition unit so as
to be reproduced by means of a monitor which acts as an image
display apparatus.
[0003] When detectors of this kind are used for medical
examinations, it is desirable to reduce the radiation dose in order
to realize a low dose whereto the patient is exposed; consequently,
only a very small amount of charge is incident on the individual
sensor elements. Therefore, the electrical charge produced in the
individual sensor elements (in dependence on the incident amount of
radiation) is also very small. This often gives rise to problems,
that is, a comparatively strong noise is superposed on the signal
read out. In this context it is known to perform so-called binning,
that is, group-wise reading out of sensor elements or the reading
out of a spatial average of pixels.
[0004] EP 0 776 124 A2 discloses a method and a device for the
imaging of an object by detection of radiation while utilizing an
image sensor system. In order to ensure an optimum ratio of image
resolution to sensitivity of the sensor, the charges of the sensor
elements are read out along the individual rows or columns of the
sensor matrix, after which they are group-wise combined. The
versions of the method proposed for this purpose are based on a
modification of control signals whose execution can be programmed.
EP 0 776 124 A2 is focused notably on the binning of CCD
sensors.
[0005] The process of reading out the charges in the individual
sensor elements or pixels in known detectors, based on conventional
matrix sensors, is conventionally performed in rows or in columns.
For example, when the first row is to be read out, the switches of
the sensors present in this row are activated by way of the
respective corresponding address line. The charges stored in the
capacitances of the sensor elements present in this row are then
drained via the respective electrical switch of the relevant sensor
elements and the corresponding read-out line.
[0006] For such sensors it is also known to use a binning operation
by extending the row-wise scans from one row to a plurality of rows
by means of known shift registers. It is a drawback, however, that
the number of combined rows or columns is constant within an
exposure.
[0007] For the column-wise reading-out, for which customarily a
charge amplifier is provided for each column and each read-out
lead, binning is achieved by interconnecting two neighboring
read-out lines in an amplifier, by interconnecting two neighboring
amplifier channels, or by calculating an average during the
subsequent image processing. According to the first two methods
reading-out can take place only column-wise or in defined groups of
columns. Even though the calculation method offers increased
flexibility in relation to the other two methods, the desired low
noise is achieved only by means of the first method; however, the
combining of each time two neighboring read-out lines concerns the
entire detector.
[0008] Considering the foregoing it is an object of the present
invention to provide a method and a sensor which offer, despite the
low incident X-ray doses, adequate quantities of charge and hence
adequate electrical signals for the formation of the image.
[0009] This object is achieved by means of a method which is
characterized in that ingoing address lines are selectably
connected, by means of individually controllable switch elements
and by way of a switching operation, to the respective next address
line in such a manner that the sensor elements of at least two
neighboring rows are activated by means of one ingoing signal, and
that corresponding outgoing read-out lines are selectably connected
to the respective next read-out line by means of individually
controllable switch elements and by way of a switching operation,
in such a manner that the charge signals read out from the
activated sensor elements of at least two neighboring columns are
combined so as to form one signal and are applied to only one
amplifier.
[0010] Under the control of switching operations either the
relevant connection of the address line or read-out line to the
switches per row or column is maintained or a connection is
established with the respective neighboring line, that is, either
the respective preceding line or the subsequent line. The
overlapping of an arbitrarily selected number of rows of activated
sensor elements as well as an arbitrarily selected number of
columns thus enables the formation of an individually programmable
binning pattern, that is, arbitrary blocks of rows and columns. Not
only the size of the individually read out surfaces of sensor
elements can thus be programmed, but also the temporal succession
of different binning patterns. Spatially inhomogeneous binning,
that is, the reading out of different surfaces in different regions
of the detector, can thus be realized.
[0011] The binning size, that is, the number of combined rows or
columns, is preferably varied within a frame or exposure. The use
of different binning sizes enables enhancement of the
signal-to-noise ratio by stronger binning in positions with a weak
signal whereas full resolution can be maintained in positions with
a good signal-to-noise ratio.
[0012] The proposed method enables variation of the size of the
binning blocks within an image. In regions in which the individual
sensor element provides only a poor signal-to-noise ratio, the
noise is reduced by binning over larger regions (2.times.2,
3.times.3, or also asymmetrically), thus enhancing the
signal-to-noise ratio. The accompanying lower spatial resolution is
accepted. In regions in which the individual sensor element already
provides a good signal-to-noise ratio, binning over only small
regions (2.times.2) or even no binning (1.times.1) is selected,
thus maintaining a high spatial resolution.
[0013] Overall, the undesirable noise can be minimized by binning
the charges read out before the amplifiers. Moreover, in conformity
with a further development of the method it should be possible to
intervene in the originally programmed read-out routine of binning
patterns during the exposure.
[0014] Such a variably programmable binning pattern is attained by
means of a sensor which is characterized in that a first unit of
switch elements is arranged between the activation means and the
electrical sensor switches and a second unit of switch elements is
arranged between the electrical sensor switches and amplifiers.
Additionally there is provided a system for the individual control
of the switch elements so as to connect, by way of a switching
operation, an arbitrary number of switch or read-out lines to each
time at least one neighboring line. Overall the control system
determines whether and which switch elements carry out a switching
operation and hence whether the corresponding sensor row or sensor
column is read out together with neighboring rows or columns or
strictly in rows or columns in the customary manner.
[0015] The switch elements in a first embodiment according to the
invention are arranged in or on the address or read-out lines and
hence in the activation or read-out circuits. In a second
embodiment they are integrated directly in the sensor matrix in
that either they are made of amorphous silicon, like the
photodiodes and the TFT switches of the sensor elements, or in that
they are arranged on (recrystallized) polycrystalline silicon.
[0016] The control system for the first and the second unit of
switch elements in a preferred embodiment is constructed as a first
and a second shift register with a plurality of shift register
elements, a respective shift register element being associated with
a switch element per address line or read-out line. The first shift
register serves as a series/parallel converter and the second shift
register as a parallel/series converter. In order to form an
arbitrary binning pattern with plane groups of sensor elements
which are interrupted by conventionally read out rows or columns,
the shift registers are constructed in such a manner that the
individual shift register elements which are not to participate in
a binning operation are automatically skipped. One clock pulse thus
suffices to address a next block of rows which is separated from a
first block by the skipped row.
[0017] The control shift register is preferably constructed so as
to be double buffered. A new binning pattern can thus be clocked
into a first storage location whereas the non-modified binning
control information for the corresponding blocks of rows or columns
is retained in a second storage location. After completion of the
block the new control information is transferred from the first
storage location to the second storage location of the control
shift register.
[0018] Instead of the shift register, or in addition thereto, a
random addressable register is used so as to realize fast updating
of the binning pattern. A double-buffered chart is again
preferred.
[0019] Further details and advantages of the invention are apparent
from the dependent Claims and the following description in which
the embodiments of the invention as shown in the Figures are
further elucidated. Therein:
[0020] FIG. 1 shows a circuit diagram of a sensor matrix with an
activation circuit influenced in accordance with the invention;
[0021] FIG. 2 shows the corresponding circuit diagram of a sensor
matrix with a read-out circuit influenced in accordance with the
invention;
[0022] FIG. 3 shows an X-ray examination apparatus which includes
an X-ray sensor in accordance with the invention.
[0023] FIG. 1 shows a circuit diagram of an X-ray sensor matrix 1
with two shift registers for activating the switches of the sensor
elements; FIG. 2 shows the circuit diagram of an X-ray sensor
matrix with two shift registers for reading out the activated
electrical switches of the sensor elements.
[0024] FIG. 1 shows a part of a sensor matrix 1 in which the sensor
elements are arranged in rows and columns. A matrix of this kind
may include, for example, 2000.times.2000 sensor elements, only a
few of which are shown by way of example in the FIGS. 1 and 2. For
a better understanding of the following description, the sensor
elements in FIG. 1 are associated with a given, consecutively
numbered row, so S.sub.1,1, S.sub.1,2 etc.; the first index
indicates the row and the second index the column.
[0025] For each pixel there is provided an X-ray sensitive sensor
element which includes a photosensor 2, a capacitance and a switch
3. The photosensors themselves may already be X-ray sensitive when
use is made of suitable semiconductors. However, a light-sensitive
photodiode may also be concerned, the conversion of X-rays into
light then taking place by means of an additional scintillator. The
photosensors are represented by the photodiodes 2 in the FIGS. 1
and 2. The electronic switches 3 are preferably constructed as TFTs
(thin-film transistors). The photosensor with scintillator converts
incident X-rays into charges which are collected by the
capacitance. The capacitance may be either an integral part of the
photodiode or be constructed as a separate component. The charge
stored in the capacitance after a given period of time is a measure
of the radiation intensity. This charge can be read out, via the
electrical switches, for each sensor element. To this end, an
address line 4 to 8 is provided for each row of the sensor matrix.
The electrical switches 3 of the associated row can be activated
via these address lines.
[0026] For each column of the matrix which is partly shown in FIG.
1 there is provided a respective read-out line 9, 10, 11, the
circuits succeeding the read-out lines being clarified in FIG. 2.
The read-out lines, denoted by the reference numerals 9 to 13 in
FIG. 2, are connected to respective amplifiers 14 to 18 which
conduct the output signals to a multiplexer 19. The multiplexer 19
delivers the electronic image signal to a monitor or an image
processor.
[0027] As opposed to the conventional sensor matrices, where
exclusively a binning pattern which is identical across the entire
detector can be selected, the sensor matrix according to the
invention enables, in conjunction with preceding and succeeding
circuits, reading out in conformity with a locally selectable or
programmable binning pattern in an image is possible, i.e. a
non-constant binning pattern.
[0028] To this end, the sensor matrix 1 is preceded by two shift
registers 20, 21 in the embodiment shown in FIG. 1. The shift
register 20 is also referred to as an activation means and consists
of a plurality of successively connected memories or shift register
elements 22 to 26, the number of which corresponds to the number of
rows of the sensor matrix. The construction of the circuit is such
that a shift pulse (in this case denoted by the reference DC for
Data Clock) causes all memories to shift their information (DI for
Data Input) to the respective neighboring memory. This shift
register is succeeded by switch elements 27 to 30. In the present
embodiment the address lines 5 to 8 comprise such a switch element;
the address line 4, however, does not have a predecessor line and
hence does not require a switch element.
[0029] The address lines either are connected directly to the
corresponding row of sensor elements or the individual address
lines are interrupted by means of the switch elements and a
connection to the relevant predecessor address line is established.
The sensor elements of at least two neighboring rows can thus be
activated by means of one ingoing signal. In the present example
the fifth line (8) is connected to the fourth address line (7), the
fourth line (7) is connected to the third address line (6), the
third line (6) is connected to the second address line (5) and the
second line (5) is connected to the first address line (4). As a
result, the information of the first address line (4) activates the
sensor elements of the rows 1 to 5.
[0030] An arbitrary number of associated rows can be activated by
arbitrary control of the switching processes of the switch
elements. The switch elements 27 to 30 are controlled by means of
an additional shift register 21. Information (CI for Control Input)
in the form of control instructions for switching over the relevant
address lines is output at the rhythm of the shift frequency (CC
for control clock).
[0031] Amplifiers 50 to 54 are inserted in the address lines
between the register and the sensor matrix. Alternatively, level
converters could be used. Such converters boost the often small
output voltages of the purely digital control signals (for example,
amounting to 5 V for the shift register) to the value required so
as to control the TFTs, which usually require from 20 to 25 V.
[0032] The switch elements 27 to 30 in the address lines or the
activation switch elements may be TFTs. Using a double construction
with opposed control, these transistors implement a respective
switch element. In this case the (analog) voltage of the activation
line could be switched behind the amplifier or level converter. In
the embodiment shown in FIG. 1 the amplifiers 50 to 54 are preceded
by the switch elements 27 to 30. Alternatively, use can be made of
a two-input digital multiplexer whose digital output signal is
raised to the necessary amplitude by means of amplifiers or level
converters.
[0033] Only purely analog switches, for example TFT or (MOS)FET,
are suitable for the switch elements 31 to 34 in the read-out
lines, because the charges are transferred from the sensor elements
without modification.
[0034] The control of switching operations of switch elements for
the read-out lines (FIG. 2) is the same as described with reference
to FIG. 1. FIG. 2 again shows the sensor matrix 1 only partly, that
is, in this case with three rows and five columns. In conformity
with the known method the charge of the sensor elements of each
column is read out by means of a read-out line 9 to 13 which is
connected to a respective amplifier, in this case denoted by the
references 14 to 18. The outputs of the individual amplifiers are
combined in a multiplexer 19 and output as an output signal. As a
further development according to the invention there is provided a
unit of switch elements 31 to 34; the second and the subsequent
columns now include a switch element which is inserted in the
relevant read-out line. The switch elements connect the relevant
read-out line either directly to the associated amplifier or to the
respective neighboring read-out line. In the present example the
second read-out line (10) can be connected to the first read-out
line (9), the third read-out line (11) can be connected to the
second read-out line (10) etc. by way of switching operations. A
collected amount of charge is thus applied to the amplifier (14)
only by the first read-out line (9).
[0035] The individual switch elements 31 to 34 are controlled by
means of a shift register 35. Control itself is analogous to the
described control of the address lines.
[0036] FIG. 3 shows a medical X-ray examination apparatus 41 which
is provided with a detector 42 which includes an X-ray sensor in
accordance with the invention. The X-ray examination apparatus 41
includes a table 43 for a patient. An X-ray source 44 is mounted
underneath the table. The X-ray detector 42 is mounted so as to be
adjustable relative to the radiation source 44. In order to form an
X-ray image, the patient is irradiated by means of an X-ray beam.
An X-ray image is formed by means of the detector in dependence on
local differences in the X-ray absorption by the patient. The X-ray
sensor converts this image into electrical signals which are
applied to the monitor 45 for displaying the X-ray image. Despite
the use of low radiation doses so as to reduce the patient load,
the sensor in accordance with the invention enables the reading out
of adequate charge intensities for the formation of the X-ray
image, reading out taking place in blocks of rows and columns of
programmable size.
* * * * *