U.S. patent application number 10/860458 was filed with the patent office on 2011-01-06 for system and method for creating a visualization indicating relationships and relevance to an entity.
This patent application is currently assigned to CERNER INNOVATION, INC.. Invention is credited to Steven Crooks, Christopher Finn, David Jr. McCallie, David O'Larte.
Application Number | 20110004529 10/860458 |
Document ID | / |
Family ID | 34397042 |
Filed Date | 2011-01-06 |
United States Patent
Application |
20110004529 |
Kind Code |
A2 |
Finn; Christopher ; et
al. |
January 6, 2011 |
SYSTEM AND METHOD FOR CREATING A VISUALIZATION INDICATING
RELATIONSHIPS AND RELEVANCE TO AN ENTITY
Abstract
A system and method for generating interactive visualizations of
clinical relationships is provided. The method includes determining
what entities have had some clinical contact with an entity of
interest and collecting data regarding each entity and each
entity's contacts with the entity of interest. The collected data
can be filtered to remove unwanted entities and properties.
Algorithms are applied to generate clinical relevance scores, which
represent the relationship between the entities and the entity of
interest. In addition, each entity is assigned components that
allow for some executable behavior when an entity is selected in a
display. Finally, the appropriate visualization is selected and a
display of the relationships is generated.
Inventors: |
Finn; Christopher; (Liberty,
MO) ; Crooks; Steven; (Gladstone, MO) ;
McCallie; David Jr.; (Stilwell, KS) ; O'Larte;
David; (Peculiar, MO) |
Correspondence
Address: |
SHOOK, HARDY & BACON L.L.P.;(Cerner Corporation)
Intellectual Property Department
2555 GRAND BOULEVARD
KANSAS CITY
MO
64108-2613
UNITED STATES
816-474-6550
816-421-5547
|
Assignee: |
CERNER INNOVATION, INC.
10975 Benson Drive Su. 550, 12 Corporate Woods
Overland Park
KS
66210
|
Prior
Publication: |
|
Document Identifier |
Publication Date |
|
US 20050076060 A1 |
April 7, 2005 |
|
|
Family ID: |
34397042 |
Appl. No.: |
10/860458 |
Filed: |
June 2, 2004 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60/509,003 |
Oct 6, 2003 |
|
|
|
60/546,935 |
Feb 23, 2004 |
|
|
|
Current U.S.
Class: |
705/20 |
Current CPC
Class: |
G16H 40/20 20180101;
G06Q 20/201 20130101 |
Class at
Publication: |
705/020 |
International
Class: |
G06F 17/00 20060101
G06F017/00; G06Q 20/00 20060101 G06Q020/00 |
Claims
1. One or more computer-storage media having computer-executable
instructions embodied thereon that, when executed, perform a method
in a computer system of generating a visualization model used to
display clinical relationships between a plurality of physical
entity elements and a physical entity of interest, the method
comprising: retrieving properties for each of the entity elements
and the entity of interest, the properties including at least one
activity type indicative of clinical contact between each of the
entity elements and the entity of interest; calculating a frequency
for the at least one activity type, wherein the frequency is a
number of clinical interactions between the entity element and the
entity of interest; calculating a clinical relevance score for each
entity element based on the at least one activity type indicative
of clinical contact between each of the entity elements and the
entity of interest, the clinical relevance score weighting certain
properties above others to rank each entity element's clinical
relevance to the entity of interest; and generating a visualization
model using the retrieved properties of each entity element and the
calculated clinical relevance scores, the model representing the
clinical relationship and clinical relevance of each entity element
relative to the entity of interest.
2. The one or more computer-storage media of claim 1, further
comprising combining the retrieved properties and sorting by entity
element after the properties are retrieved.
3. The one or more computer-storage media of claim 1, further
comprising filtering the retrieved properties to eliminate any
properties that are not of interest.
4. The one or more computer-storage media of claim 1, further
comprising repeating the retrieving, calculating and generating
steps for each of a plurality of time periods.
5. The one or more computer-storage media of claim 1, wherein the
relationships are clinical relationships, the entity of interest is
an individual receiving healthcare and the entity elements are
individuals providing the healthcare or resources for providing
healthcare.
6. The one or more computer-storage media of claim 1, further
comprising displaying, on a display device, the visualization of
the relationships using the generated visualization model.
7. The one or more computer-storage media of claim 6, further
comprising assigning icons to each of the displayed entity
elements.
8. The one or more computer-storage media of claim 7, wherein the
size, shape and color of the assigned icons are used to convey
relevance and property information.
9. The one or more computer-storage media of claim 7, further
comprising associating at least one executable action with at least
one of the entity elements.
10. The one or more computer-storage media of claim 7, wherein the
executable action is an executable uniform resource locator (URL)
that is executed by a mouse click on the entity icon.
11. The one or more computer-storage media of claim 1, wherein the
contact between the entity element and the entity of interest is an
actual physical contact.
12. The one or more computer-storage media of claim 1, wherein the
contact between the entity element and the entity of interest is a
request that will direct a physical contact between an entity
element and the entity of interest.
13. The one or more computer-storage media of claim 1, wherein the
contact between the entity element and the entity of interest is a
passive contact between the entity element and the entity of
interest.
14. A computer system for generating an entity-of-interest focused
visualization to display clinical relationships between an entity
of interest and a plurality of entity elements, along with a
clinical relevance of the entity elements to the entity of
interest, the system comprising: at least one originating system
adapter component that retrieves properties of entity elements that
have had clinical contact with the entity of interest from an
originating system, the properties including at least one activity
type indicative of clinical contact between the entity element and
the entity of interest; a visualization model generation component
that calculates a clinical relevance score for each entity element
based on a clinical relevance algorithm and the at least one
activity type indicative of contact between each entity element and
the entity of interest; and a visualization engine that generates a
visualization using the retrieved properties of each entity element
and the clinical relevance score calculated for each entity
element.
15. The computer system of claim 14, further comprising a filter
component that operates to remove unwanted properties from the
retrieved properties prior to the calculation of the relevance
score.
16. The computer system of claim 14, wherein a plurality of
originating systems are available, further comprising an
originating system adapter for each available originating
system.
17. The computer system of claim 14, wherein the visualization
engine also assigns iconography to the entity elements to be
displayed and associates action classes with each entity
element.
18. The computer system of claim 14, wherein the contact between
the entity elements and the entity of interest is indirect
contact.
19. A user interface embodied on a computer readable medium and
executable on a computer, comprising: an entity of interest icon;
and a plurality of entity element icons, each representing an
entity element associated with the entity of interest, wherein each
entity element icon has associated therewith an executable action,
and wherein the arrangement of the entity element icons relative to
the entity of interest icon conveys information about a clinical
relationship between the entity of interest and each entity element
based on one or more clinical contacts between the entity of
interest and each entity element such that a more relevant entity
element icon is arranged closer to the entity of interest icon than
a less relevant entity element icon or the more relevant entity
element icon is larger than the less relevant entity element
icon.
20. The user interface of claim 19, wherein the relationship
information is conveyed using the color, shape and positioning of
the entity element icons.
21. The user interface of claim 20, wherein the executable action
is a uniform resource locator (URL) unique to the entity element
with which it is associated.
22. The user interface of claim 21, wherein the entity element
icons represent healthcare resources and wherein the entity of
interest icon represents a person receiving healthcare.
23. One or more computer-storage media having computer-executable
instructions embodied thereon that, when executed, perform a method
in a computer system of generating a visualization model used to
display clinical relationships between a plurality of physical
entity elements and a patient, the method comprising: retrieving
properties for each of the entity elements relative to the patient
for a plurality of time periods, the properties associated with one
or more physical clinical contacts between each entity element and
the patient, wherein the plurality of time periods is a
predetermined length of time including a range window and an
overall time period for which to calculate a frequency of the one
or more physical clinical contacts between each entity element and
the patient, wherein the frequency is a number of clinical
interactions between each entity element and the patient;
generating a visualization model representing the clinical
relationship and a clinical relevance of each entity element
relative to the patient using the retrieved properties of each
entity element for each time period, wherein the clinical relevance
is calculated based on the retrieved properties and the frequency
of the one or more physical clinical contacts between each entity
element and the patient; and sequentially displaying the clinical
relationships using the generated visualization model for each time
period.
24. The one or more computer-storage media of claim 23, wherein the
clinical relevance of each entity element relative to the patient
is determined by calculating a relevance score for each entity
element using the retrieved properties of each entity element, the
relevance score weighting certain properties above others to rank
each entity element's relevance to the patient.
25. The one or more computer-storage media of claim 24, further
comprising filtering the retrieved properties to eliminate any
properties that are not of interest.
26. The one or more computer-storage media of claim 23, wherein the
relationships are clinical relationships, the patient is an
individual receiving healthcare, and the entity elements are
individuals providing the healthcare or resources for providing
healthcare.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. Provisional
Patent Application Ser. No. 60/509,003, filed Oct. 6, 2003 and U.S.
Provisional Patent Application Ser. No. 60/546,935 filed Feb. 23,
2004.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0002] Not applicable.
TECHNICAL FIELD
[0003] The present invention relates generally to the field of
computer software. More particularly, the present invention relates
to a computerized system and method for creating interactive
visualizations of all entities that have established some form of
contact or relationship with an entity of interest.
BACKGROUND OF THE INVENTION
[0004] When a person enters a health system, he or she will
frequently come in contact with dozens of providers, clinicians,
staff members and other patients or persons during the course of
care. Healthcare providers often have a need to determine, during
this course of care, the care the person has received and who
provided that care. In the healthcare industry, this information
can be recorded on various forms manually, or can be housed within
a computerized clinical information system. Various people within
the healthcare system may also need to determine which person has
the most relevant information about a particular patient. However,
it may be difficult to determine "who knows the patient best." Such
information can be gleaned by going through records individually,
getting detail about each encounter, and making an estimation of
who to contact. However, this process is time consuming, even in an
electronic system, and may yield information that ultimately is
useless. For example, knowing that a particular nurse is most
familiar with a patient may not help solve the immediate need if
that nurse is not present or otherwise available.
[0005] Likewise, infection control issues may arise in a healthcare
setting. In such situations, it is important to identify the
persons and resources (rooms, equipment, etc.) that have been
exposed to a particular patient. It is also important to know who
has subsequently come into contact with those individuals or
resources who were exposed to the patient, and the individuals and
resources who contacted those individuals and resources exposed to
the patient. Again, the information can often be recreated from
manual or electronic records, but the process is time consuming and
inefficient.
[0006] There exists, therefore, a need to quickly and efficiently
obtain this kind of information. The information that is needed
should also be displayed in a manner that quickly conveys the most
relevant information. In addition, it would be advantageous if the
display is dynamic and interactive. Most traditional visualizations
in the healthcare setting show retrospective views of data
specifically stored about the person of interest, often missing the
value of dynamic information about relationships that are relevant
to the person. However, these visualizations (and the tools
associated with these visualizations) do not facilitate real-time,
interactive application management during the care process.
[0007] Accordingly, there is a need for an efficient system and
method for determining the relationship between an entity of
interest and the various entities that have had contact with that
entity, such as a clinical relationship between a patient and those
that have contact with the patient. There is also a need for a
system and method for displaying clinical relationship information
that allows a user to perform a variety of functions such as
viewing and analyzing relationships, initiating collaborative work,
and carrying out infection control.
SUMMARY OF THE INVENTION
[0008] The present invention generally relates to a system and
method for generating interactive visualizations of clinical
relationships. The method includes determining entities that have
had some clinically-related contact with an entity of interest and
collecting data regarding each entity and each entity's contacts
with the entity of interest. The collected data can be filtered to
remove unwanted entities and properties. Algorithms are applied to
generate clinical relevance scores, which represent the
relationship between the entities and the entity of interest. In
addition, each entity is assigned components that allow for some
executable behavior when an entity is selected in a display. Also,
the appropriate visualization is selected and a display is
generated.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
[0009] The present invention is described in detail below with
reference to the attached drawing figures, wherein:
[0010] FIG. 1 is a block diagram of a computing system environment
suitable for use in implementing the present invention;
[0011] FIG. 2 is a flowchart representative of a computer program
for generating and displaying a view of clinical relationships;
[0012] FIG. 3 is a flowchart detailing the steps carried out by the
visualization engine component of the present invention; and
[0013] FIGS. 4-6 are screen shots illustrating exemplary
visualizations generated by the system.
DETAILED DESCRIPTION OF THE INVENTION
[0014] The present invention is directed to a system and method for
creating visualizations representing relationships between various
entity elements and an entity of interest. The entity of interest
is the focus of the visualization or display. In other words, it is
the entity for which care information is generated and visualized.
The entity of interest can be a person, such as a patient, a
provider, or a caregiver, or can be any inanimate entity, such as a
resource (a location, a bed, a room, a piece of equipment, etc.) or
other entity. Entity elements are those physical entities that have
had some clinical relationship or contact with the entity of
interest. An entity element can also be any type of physical
entity, either animate or inanimate such as a person or resource
capable of having a clinically-related contact with an entity of
interest. In one example, discussed below, the entity of interest
is a patient in a healthcare setting, and the entity elements are
individuals providing various levels of care to that patient. It
should be understood, however, that this example is not meant to be
limiting and that the invention is applicable to a broad range of
entities of interest and entity elements. For instance, a role
associated with a class of people is a possible entity element. In
this example, a number of different nurses may represented as a
single entity element for the nurse role. In a related alternative,
a type of resource is a possible entity interest. In this example,
a number of x-ray machines may be represented as a single entity
element. This is desirable in many clinical situations because the
type of resource is oftentimes more relevant to the provision of
care than the specific resource such as in the example of an x-ray
machine.
[0015] FIG. 1 depicts a block diagram of an exemplary computer
system 100 for generating and displaying clinical relationships in
accordance with the present invention. The system 100 includes
various originating systems 102 which provide information to
originating system adapters 104. Originating systems are systems
that contain potential data used to generate the visualizations
discussed below. For example, an originating system could be a
clinical data repository, an e-mail system, a presence tracking
system, or a financial system. An originating system adapter 104 is
provided for each originating system 102 of potential use. A
visualization model generation component 106 communicates with the
originating system adapters 104. Component 106 takes the data
collected by the originating system adapters 104 and manipulates
the data using various filters and algorithms as will be described
in more detail below. As will also be described below, component
106 generates the collection of information needed to render the
desired visualization or display. A visualization engine 108 takes
the visualization model generated by component 106 and generates
the appropriate visualization for display 110.
[0016] The use of the components of FIG. 1 is described in more
detail with reference to FIG. 2. The overall method is discussed
first and will be followed by an example. The method begins when
originating system adapters 104 are invoked to retrieve data from
various originating systems 102, as shown at 202. The data
retrieved by the adapters 104 are a set of properties (or features)
for a given set of entities. This data includes the various entity
elements that have had contact (or "touches") with the entity of
interest and each entity element's associated properties entity.
Entity properties collected from originating systems 102 by
adapters 104 may include such things as the entity's name, a unique
entity identifier, the primary role of the entity (e.g., attending
physician, nurse practitioner, etc.), activity type data, presence
status, and generally any other attribute that may be of
interest.
[0017] Activity types are the data elements indicative of the
contact between an entity element and the entity of interest. A
contact may be an actual physical contact between an entity element
and the entity of interest, a request that will direct a physical
contact between an entity element and the entity of interest or a
passive contact between the entity of interest and the entity of
interest. Examples of actual (or direct) physical contacts include
administering a medication to a patient, performing a procedure on
a patient and common physical presence of the entity element and
entity of interest. If the entity element is a room, piece of
equipment or type of equipment, or similar resource, examples of
actual physical contact include using a resource for a test or
admitting a patient to the resource. Examples of requests that will
direct a physical contact between an entity element and the entity
of interest include ordering a medication or test, or scheduling a
room. Examples of passive (or indirect virtual) contacts include
opening a patient's chart, adding a problem to the list of problems
for the patient and reviewing an image of the patient.
[0018] The presence status property may indicate contact-type
information, such as the entity element's current location,
online/offline status, availability, and/or device
capabilities.
[0019] As shown at 204, the retrieved entity elements and
properties collected at 202 are combined into a single list and
grouped together, or sorted, by entity element. Next, exclusionary
filtering is applied to the generated list to remove undesired
entities and properties, as shown at 206. In the following step,
frequencies are calculated for each activity type or event of
interest and are summarized into either a single frame or into
multiple frames, as shown at 208. A single frame is used to display
a static or "snapshot" view. Multiple frames are used to display a
dynamic or "animated" view that can be used to show changes over
time, or trends. The frequency information is used in subsequent
steps as described below. This frequency information indicates the
number of times a particular activity type is encountered. For
example, the frequency information can be instructive regarding the
number of occasions a particular entity, such as an attending
physician, opens the patient's chart. The number of days an entity
element interacted with the element of interest is also calculated
at this stage. This is sometimes referred to as the number of days
"touched." For a particular entity element, such as a doctor, the
days touched reflects the number of days the doctor had some
interaction with the element of interest, such as the patient.
[0020] The next step involves applying clinical relevance
algorithms to the data collected to calculate clinical relevance
scores for relevance perspectives, as shown at 210. A relevance
perspective defines a mathematical function modeling a particular
use case. For example, one relevance perspective identifies the
most relevant clinical decision maker for a patient. By way of
example, a mathematical function for this relevance perspective
places a higher value on the number of medication orders or
invasiveness of the procedures. Another relevant perspective
identifies the caregiver who knows the patient best. A mathematical
function for this relevance perspective places a higher value on
the number of touches rather than the type of touch. Generally, a
clinical relevance algorithm applies a mathematical function
against the data collected about particular activity types,
including the frequency of such activities, in order to calculate a
clinical relevance score. The mathematical function may be a simple
linear function or a more complex function applied to the activity
type data. The relevance perspectives can either be defaulted by an
application for a particular use case, selected by a user from a
predefined set of relevance perspectives, or defined by a user by
controlling the nature (e.g., weights, constants, terms) of the
mathematical function applied against the activity data and
activity type data. This step generates a number of relevance
perspective vectors, which each contain information about an entity
element (including its properties) and a clinical relevance score
for a particular relevance perspective. A more specific explanation
of the relevance score calculation is set forth in the example
below.
[0021] As shown at 212, at least one executable action is next
associated with each entity element to allow for interactive views.
The executable action may be assigned to each entity element. The
executable action defines a behavior that occurs when the entity
element is acted upon by the user. For example, an executable
action could invoke a particular Uniform Resource Locator (URL) to
begin a real-time collaboration session with the user. An
executable action could also be associated with a Java class file
that, when invoked, changed database information associated with
the entity element. Likewise, an executable action could be a
JavaScript function which, when invoked by the user, changes the
nature of the relationship between the entity element and the
entity of interest. Following this step, the visualization model
for the display is complete. The visualization model includes the
collection of relevance perspective vectors, the executable actions
associated with entity elements, the associated frames (for
animation), and other information necessary for display (either
static or animated).
[0022] After a visualization model has been generated, the
visualization engine 108 generates the display. Different
visualization adapters within engine 108 are selected, as shown at
214, and invoked to render the appropriate display, as shown at
216. The visualization engine 108 and generation of visualizations
will be described in more detail below.
[0023] A wide variety of displays can be generated by the present
invention. Because the visualization model and the visualization
displays are separate, the displays may be a graphical view or a
tabular view. In addition, the displays or visualizations can be
separated into families of use cases. They can be either
point-in-time (snapshot) or continuous time (animated). Snapshot
views typically have lower-stress data-retrieval requirements and
are designed to help users make quick decisions visually through
rapid assimilation of visual data. Animated views are designed to
show real-time status or trends, and typically will have more
demanding data-retrieval requirements.
Exemplary Display Generation
[0024] The following simplified textual example illustrates the
method of the present invention showing how entity element
information is accumulated and turned into a visualization or
display. For clarity, this example shows the generation of a
snapshot view, with commentary at appropriate times for variations
for multiple frames.
[0025] Under the first step shown at 202, originating system
adapters 104 collect entity element information and their
properties from originating systems 102. In this example, the
entity of interest is a patient named "Charles Kramer," and the
user wishes to see a display of "online entities displayed by
clinical relevance and recency of touch." A clinical information
system adapter 104 provides a set of activities related to clinical
relevance. For illustrative purposes, assume the activities include
the number of orders placed by an entity element against Charles
Kramer (MedOrder), documents authored or edited by an entity
element for Charles Kramer (DocAuthored), and number of times the
chart was opened by a particular entity element for Charles Kramer
(ChartOpen). Since the system in this example is to display "online
entities," the system will also filter based on presence status so
a real-time collaboration system adapter 104 is also needed.
[0026] These two adapters 104 provide raw information, which is
then grouped together by entity element as shown at 204, with an
excerpt shown in Table 1. Table 1 displays the entity elements and
a unique identifier associated with each entity element. It also
shows the activity type, date and a property value, if applicable.
TABLE-US-00001 TABLE 1 Excerpt of the "Touch Collection" following
the Aggregation Phase. Property Entity Element Unique Id Activity
Type Date Value Paige Kaeding PK12345-123 ChartOpen 20030101-1715
Paige Kaeding PK12345-123 ChartOpen 20030102-1915 Paige Kaeding
PK12345-123 ChartOpen 20030103-2037 Paige Kaeding PK12345-123
ChartOpen 20030105-1843 Paige Kaeding PK12345-123 ChartOpen
20030105-2304 Paige Kaeding PK12345-123 DocAuthored 20030101-1720
Paige Kaeding PK12345-123 DocAuthored 20030102-1935 Paige Kaeding
PK12345-123 DocAuthored 20030103-1900 Paige Kaeding PK12345-123
DocAuthored 20030105-2330 Paige Kaeding PK12345-123 Presence
20030106-0005 Online Paige Kaeding PK12345-123 Role 20030106-0005
R.N. J Billings JB56789-435 MedOrder 20030101-1212 J Billings
JB56789-435 MedOrder 20030101-1212 J Billings JB56789-435 ChartOpen
20030102-1245 J Billings JB56789-435 DocAuthored 20030102-1300 J
Billings JB56789-435 Presence 20030106-0005 Online J Billings
JB56789-435 Role 20030106-0005 Admitting MD Leah Phelan LP45455-909
ChartOpen 20030104-1414 Leah Phelan LP45455-909 Presence
20030106-0005 Offline Leah Phelan LP45455-909 Role 20030106-0005
R.N. . . .
[0027] If the user in this example does not care about entity
elements who are offline, the system can filter those out
immediately, as shown at 206. Note that the system may also filter
out data in the adapters 104, rather than after the list is
generated in step 204. Table 2 shows an excerpt of the combined
data after Leah Phelan (the offline entity element) has been
filtered out based on the "Property Value" associated with the
Presence activity type. After this filtering stage, "Property
Value" information can be stored directly with the entity element.
TABLE-US-00002 TABLE 2 Elements and Associated Properties after
Exclusionary Filtering. Entity Element Unique Id Activity Type Date
Property Value Paige Kaeding PK12345-123 ChartOpen 20030101-1715
Paige Kaeding PK12345-123 ChartOpen 20030102-1915 Paige Kaeding
PK12345-123 ChartOpen 20030103-2037 Paige Kaeding PK12345-123
ChartOpen 20030105-1843 Paige Kaeding PK12345-123 ChartOpen
20030105-2304 Paige Kaeding PK12345-123 DocAuthored 20030101-1720
Paige Kaeding PK12345-123 DocAuthored 20030102-1935 Paige Kaeding
PK12345-123 DocAuthored 20030103-1900 Paige Kaeding PK12345-123
DocAuthored 20030105-2330 Paige Kaeding PK12345-123 Presence
20030106-0005 Online Paige Kaeding PK12345-123 Role 20030106-0005
R.N. J Billings JB56789-435 MedOrder 20030101-1212 J Billings
JB56789-435 MedOrder 20030101-1212 J Billings JB56789-435 ChartOpen
20030102-1245 J Billings JB56789-435 DocAuthored 20030102-1300 J
Billings JB56789-435 Presence 20030106-0005 Online J Billings
JB56789-435 Role 20030106-0005 Admitting MD . . .
[0028] During the next step shown at 208, frequencies of the
relevant activities are calculated and summarized into either a
single frame (for a static, snapshot view) or into multiple frames
(for an animated view). Since the weighting algorithms used in step
210 in this example use the number of days an entity element
interacted with the element of interest, the system in this example
also calculates a count of "DaysTouched." Using the example data
above for a snapshot view, Table 3 shows the calculated
frequencies. TABLE-US-00003 TABLE 3 Calculated Frequencies Entity
Element Unique Id Activity Type LastDate Frequency Paige Kaeding
PK12345- ChartOpen 20030105-2304 5 123 Paige Kaeding PK12345-
DocAuthored 20030105-2330 4 123 Paige Kaeding PK12345- DaysTouched
20030105-2330 4 123 J Billings JB56789- MedOrder 20030101-1212 2
435 J Billings JB56789- ChartOpen 20030102-1245 1 435 J Billings
JB56789- DocAuthored 20030102-1245 1 435 J Billings JB56789-
DaysTouched 20030102-1245 2 435 . . .
[0029] Role information and Presence information, being properties
uniquely assigned to an individual, are not included in the
frequencies. The system and method of the present invention may be
used as an interactive practice tool. Additionally, the system and
method may be used to view clinical and social relationship
information over time through an animated view mode. For an
animation view of a particular overall time period, these
frequencies would be counted for a particular range of events
(called a range window), corresponding to a set time period (e.g.,
a sliding window of forty-eight hours) or corresponding to a set
number of activity events (e.g., the last fifty touches).
Frequencies are then subsequently recounted for a sliding range
window corresponding to an advancement of the range window by a
"step" size, indicating by how much time (or by how many activity
events or touches) the range window should be advanced before
recalculation. The range window and step size parameters are
defined as part of the visualization. For example, a range window
may be defined as "all touches occurring in a given twenty-four
hour time period", with a step size set to twenty-four hours, and
the overall time period being the last thirty days. This would
produce thirty frames, corresponding to a sequence of twenty-four
hour range window snapshots calculated every twenty-four hours for
the past thirty days. The animation view of the frames may
illustrate changes in the nature of the relationship over time.
[0030] In the next step shown at 210, clinical relevance scores are
calculated to generate relevance perspectives. In general, for this
example, the clinical relevance algorithm is: [0031] SUM
(frequency_of feature[i]*weight_of_feature[i])*(scaling_factor):
where scaling_factor normalizes the weighted sum to a given
patient-specific context (e.g., DaysTouched/TotalDays).
[0032] Depending on the particular use, the type of activity types
used in the algorithm and the weight applied to each activity type
can be adjusted. For example, if a user wants to determine who has
the most interaction with the patient, the frequency of touches is
more important than the nature of touches, and the weights applied
to particular activity types will be roughly the same. If the user
wants to determine the decision maker with the most responsibility
for the patient's health status, the nature of touches becomes
important, and the more direct and invasive activity types are
given higher weights than indirect and passive active types.
[0033] In this example, if the user is interested in identifying
the key decision maker, the visualization filter rule of interest
is "decision making" relevance. In addition, if the user is
interested in the how recently an entity element has had contact
with the entity of interest, the "recency of touch" is relevant.
The "decision making" relevance filter is frequency based, while
"recency of touch" is a simple time-based visualization rule. The
system loads the "decision making" relevance filter weights and
applies the following weighting factors: [0034] MedOrder=25,
ChartOpen=2, DocAuthored=3.
[0035] In addition, "days touched" is used to scale these values
based on the total days of interest (for this example, Jan. 1, 2003
through Jan. 6, 2003, or 6 days.) Because "recency of touch" is a
simple rule, there are no weights associated with it. So the system
simply determines the most recent event and passes the time stamp
along. Since only one frame is desired in this example, the system
can generate the clinical relationship scores and relevance
perspectives vectors by applying the weights to the collected
statistics as follows with the results shown in Table 4. [0036]
Paige Kaeding scores: [(0*25)+(5*2)+(4*3)]*(4/6)=14.67 [0037] J
Billings scores: [(2*25)+(1*2)+(1*3)]*(2/6)=18.33
[0038] TABLE-US-00004 TABLE 4 Relevance Perspectives Entity Element
Unique Id Relevance Class Socre RecencyOfTouch Paige Kaeding
PK12345-123 DecisionMaking 14.67 20030105-2330 J Billings
JB56789-435 DecisionMaking 18.33 20030102-1245 Once again, for an
animation view, these calculations would be made on a per-frame
basis, respecting the epoch and step values mentioned earlier.
[0039] Next, as shown at 212, once the URL is invoked, the
visualization model is complete and includes the following
information: the entities that will be displayed; the relevance
classes that will be displayed; the appropriate clinical
relationship scores to be displayed for each relevance class;
additional property information about the entity elements; and
frames that can fulfill the visualization of animated presentations
of the data.
[0040] In this example, the entity elements are Paige Kaeding and J
Billings, the relevance perspectives are "decision making" and
"recency of touch," clinical reference values have been generated
for each, and the additional property information includes role and
presence status.
[0041] As shown at 214, the next step involves instantiating the
appropriate visualization adapter in engine 108 (e.g., "Graph Mode
for Scalable Vector Graphics or SVG", or "table mode for HTML").
Based on the context of the request (e.g., from a web page
requiring an interactive applet, from a Java Server Page requiring
a simple PNG graphic file), different visualization adapters may be
invoked to render a display appropriate for the venue, as would be
understood by one of skill in the art.
[0042] FIG. 3 gives a detailed flow of the operation of
visualization engine 108. This is a general method of how one might
implement a visualization adapter in engine 108. Depending on the
venue, certain steps may not be appropriate. For example, if a
tabular view is used, it may be unnecessary to associate
iconography with each entity element.
[0043] As shown at 304, the process of the visualization engine 108
begins when the visualization model 302 is taken and iconography is
assigned to each entity element. The iconography assigned to each
entity element can be based on any entity property such as role
and/or presence status. Next, relevance classes are mapped to a
visual layout representation, as shown at 306. For example,
"DecisionMaking" relevance could be mapped by the size of the icon,
"RecencyOfTouch" could be mapped by the proximity of the entity
element node to the entity of interest node, and the Presence
status could be mapped by the color of the entity element node. For
size and positional mapped relevance classes, the entity element
nodes can be scaled to size for the target area that is available,
as shown at 308. This step may be required to ensure that the
entity element nodes fit within the given target area or to ensure
that the relative scale is maintained between the various entity
element nodes. Next, as shown at 310, visualization-specific layout
rules are applied. Under this step, geometric positioning is
applied to minimize the overlap of entity element nodes on the
screen or the entity elements are sorted for a tabular display.
Executable actions are then associated with each entity element, as
shown at 312. For animated views, steps 308 through 312 are
repeated for each frame to be generated, as shown at 314.
[0044] FIGS. 4-5 are examples of schematic screen shots of the
generated visualizations. FIG. 4 illustrates one graphical
visualization based on the previous example. In FIG. 4, each entity
element is represented as an entity element node 404 surrounding
the entity of interest node 402. The shape of each entity element
node represents the entity element's role. For example, each doctor
is represented as a triangle. The size of the entity element node
represents the "DecisionMaking" relevance. Here, Jonathan Billings
and Paige Kaeding are of higher relevance in decision making. The
color or shading of the entity element node represents the presence
status of the entity element. For example, red could indicate that
the entity element is unavailable, yellow could indicate that the
entity element is busy, and green could indicate that the entity
element is available. Although shown with different shading, color
is preferred. The length of the line between the entity of interest
and each entity represents the "recency of touch." Each of these
parameters (i.e. shape, size, color, length of line) can be
modified to represent a different entity element property or
relevance class.
[0045] FIG. 5 illustrates an alternative graphical form of display
for the preceding example. The location of the entity element
represents "recency of touch." In this exemplary screen shot, the
entities with the more recent touches are located toward the
upper-left corner. The size of each entity element's box represents
the "decision making" relevance. The color or shading of each
entity element's box represents each entity element's presence
status. Again, although shown with different shading, color is
preferred. As in FIG. 4, each of the parameters (i.e. location,
size, and color) can be modified to represent a different entity
element property or relevance class.
[0046] FIG. 6 illustrates another visualization based on the
previous example. In this screen shot, the entity elements are
grouped together in a tabular form. The Presence values, the
clinical relevance values, and the values for the "recency of
touch" are included in the table. The names are preferably sorted
in order of clinical relevance and are actionable. Additional
variations are also contemplated. By way of example, the Presence
values may be represented by iconography associated with the
names.
[0047] The example given above, as well as the displays shown in
FIGS. 4-6, are merely exemplary. It should be understood that these
examples are not meant in any way to be limiting. A myriad of
visualization displays and screen shots are possible utilizing the
visualization model, and are all contemplated by the present
invention. It can be seen that the visualization displays created
and shown in FIGS. 4-6 quickly and efficiently illustrate clinical
relationship information. Moreover, when the action classes are
associated with the iconography, a useful interactive display is
created. For example, executable actions can be used to allow a
user to select particular icons shown in FIG. 4 to initiate a real
time collaboration session with desired entities.
[0048] Those skilled in the art will appreciate that the present
invention represents a significant advancement of the state of the
art beyond existing applications, especially with respect to the
practical benefits it provides. The present invention provides a
logical model and means of display for clinically-relevant
relationships gleaned from one or more sources. The present
invention also provides a mechanism for weighting the parameters of
clinical relevance to change the display to highlight clinical
relevance as defined for difference purposes. In addition, the
present invention includes an interactive graphical interface that
makes the visualization capable of being used as a launch point for
various application activities ranging from specific things, such
as initiating a collaboration session with the most appropriate
people, to navigating the clinical network. Furthermore, the model
and display are capable of being merged and intersected, meaning
that it is possible to get a broader view by combining multiple
models and displays together into a single, large display. Thus,
the system can be used for a variety of functions including
initiating collaboration, managing person-personnel security and
care team relationships, and for infection control purposes.
[0049] Alternative embodiments and implementations of the present
invention will become apparent to those skilled in the art to which
it pertains upon review of the specification, including the drawing
figures. For example, any of a number of relevance algorithms may
be employed to generate a relevance score. Accordingly, the scope
of the present invention is defined by the appended claims rather
than the foregoing description.
* * * * *