Teaching:TUW - UE InfoVis WS 2007/08 - Gruppe 07 - Aufgabe 4: Difference between revisions

From InfoVis:Wiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
(40 intermediate revisions by 4 users not shown)
Line 6: Line 6:
[[Image:20070920_blutuntersuchung.jpg]]
[[Image:20070920_blutuntersuchung.jpg]]


== Analysis of the application area ==
== Description of the application field ==
Our sample data belongs into the medical field of application. We are dealing with diagnostic findings which are basically sets of multiple tests that play a supportive role in doctors' decisions about the diagnosis. Given these two examples we can see that the second finding holds multiple times as much data as the first one. This indicates that each patient has a specialized set if tests that has been chosen for him by his doctor.
The two tables above show medical data, which are two diagnostic findings of a patient recorded at different dates. In each table row, one examination, it's result, the unit of the result and the interval of "optimal" values are shown. The examination results descripe the amount of chemical compounds or substances in the patient's body. Diagnostic findings play an important role for medical diagnosis.<br>
The first table consists of more rows than the second. Perhaps less substances in body were examined or the tableview shows only  a short section of the results.<br>


The data has been arranged into four columns. The first column holds the name of the test or the name of the subject the patient has been tested on. The second column holds the outcome of the test. In all cases we can see here, we are dealing with numbers. But some of these numbers are floating point numbers, others are integers. One of these tests even returns multiple values as result. Some of the cells in this column have been marked using the star character (*). This indicates that a value does not lie in an interval specified in the fourth column. The third column specifies the units of the results. The fourth column specifies the reference values for each of these tests, i.e. an interval that specifies acceptable values for this test. Multiple forms of intervals are being used, e.g:
Each table has four columns. The names of the examinations and the name of the substances the patient has been tested on, respectively, are written in the first column. The result values - numerical values - for each examination are denoted in the second column. Some values are floating point numbers, others are integers. The reason of using floating numbers is, that small variances of the values may cause high differences in diagnosis. One examination has more than one result ("Differential Blutb."), but we don't know the meaning of this results. Some of the cells in this column are marked with "*". This indicates that a value does not reside in an interval (specified in the fourth column). In the third column, the units of the results are shown. The fourth column lists the intervals of reference values - the "optimal" values - for each examination. If a value resides within an interval, the result is "good" - which means that the patient has no lack and no overrun of a specific element. <br>
Multiple forms of intervals are being used, e.g:
* bis 200
* bis 200
* 220-480
* 220-480
Line 19: Line 21:


=== Target group ===
=== Target group ===
The target group (the data has been produced for) is definitely hospital/medical staff. We think that the most common people can't start anything with this data. They can basically determine which test results don't match the reference interval and therefore didn't go well but this data needs to be interpreted by a specialist.
The target group (the data has been produced for) is definitely hospital/medical staff. We think that people who do not have any medical knowledge, are not able to make a diagnosis based on the data. They can basically determine which test results do not reside within the reference intervals. The data is very private and should only be seen from people authorized people - like medical stuff. Of course the application can also be used to give the patient an overview of his diagnosis data. For that purpose it is important that the attending physician is at present and can present the visualised data with several views and funded knowledge. For that reason we try to display the given data as intuitive as possible to ease the handling of the appication for the medical stuff aswell as patients that are assisted by a physican should be able to understand basic coherences and can therefore gain a better understanding for ongoing inquiries.
Another thing is that this data is private and should therefore not be seen by anyone except for the doctor and the patient.  


== The DIAMONT application ==
== Concept & Mockup for: The DIAMONT application ==
We named our application '''DIAMONT''' which stands for '''DIA'''gnosis '''MON'''itoring '''T'''ool. This tool has been designed to be used by the hospital's medical staff including doctors and nurses. It shall provide a more easier and better organized view on the data from test sets. It is meant to be realized either in a form of a desktop application or as a PDA (handheld-gadget) application. The accessed data has to been accessible remotely, therefore needs to be located on a server.
We named our application '''DIAMONT''' which stands for '''DIA'''gnosis '''MON'''itoring '''T'''ool. This tool has been designed for medical staff, i.e. doctors and nurses. It shall provide a more easier legible and better organized view on the data from the test sets. The application is made as a desktop application or as a PDA (handheld-gadget) application. The idea is, to share the data with all hospitals and medical institutions. Therefore, the software has to be installed in each medical institution and the data itself is located on a remote server.  


The next subsections will provide a detailed look at the interface of the DIAMONT application. The reference numbers reference the red numbers located on the scrrenshot of the application.
In the following, we will provide a detailed description of DIAMONT's interface. The red colored numbers mark the parts of the interface we want to describe in detail.


[[{{ns:6}}:0708_Gruppe_07_diamont.jpg]]
[[{{ns:6}}:Diamont_screen1.jpg]]<br>
<b>Figure 1</b> <br><br>
[[{{ns:6}}:Diamont_screen2.jpg]]<br>
<b>Figure 2:</b> point marked with black circle can be edited in the medical notes sheet (see Figure 3)<br><br>
[[{{ns:6}}:Diamont_screen3.jpg]]<br>
<b>Figure 3</b><br><br>


=== ad 1: Patient Search ===
<b>ad 1: Patient Search</b><br>
 
The patient search section allows to search for specific patients. By simply typing in a name (or other inforamtion which can be used to identify a patient) and pressing the "Suche Patienten"-Button, the application displays a list of records in the main window. By single-clicking a row, more detailed information about the selected patient is shown in the main window, which is for this purpose an "Patient view" window. Data of the patient can be: name, date of birth, dates of hospitals stays etc. Double-clicking a row will load the patients test values to the list of available tests (see 3).
Jakob
<br><br>
 
<b>ad 2: Patient information</b><br>
=== ad 2: Patient detail information ===
This section displays the patient's name and birth-date. More detailed information about the patient is given (in the main-window) when the name of the patient is clicked with the mouse. The information is shown then in the main window and "Patiens view", respectively. There, the patient's data can be modified.
Jakob
<br><br>
 
<b>ad 3: List of available tests</b><br>
=== ad 3: List of available tests ===
The list of available tests shows the names of the tests/examinations that are represented in the dataset. By dragging a test into the main window (see 4), the examination information is visualised in the main window. Values that reside below or above the optimum range - at the chosen date - (see 6) are highlighted red. Values that are within the reference range are black. This feature allows the user to quickly scan the tests for outstanding values, that might be interesting for further research.
Jakob
<br><br>
 
<b>ad 4: Main Window </b><br>
=== ad 4: Main Window ===
The user can drag and drop one or more tests (the amount is not limited) from the list (see 3) into the main window and the sub-windows, respectively. Empty sub-windows (where no examinations are dropped)show a message that says, that an examination can be dragged and dropped into this window. Each sub-window, which contains an examination, displays detailed information of the tests/examinations over time. The values obtained over time are visualised with a graph (line chart). The sub-windows are divided by horizontal lines into three parts. Examination values that reside above the reference interval are drawn at the top. The part in the middle shows values that reside within the interval. The bottom part shows test results that reside below the reference interval. Additionally, the values are marked with colored circles: red, if the value is too low or too high; green, if the value resides within the optimal interval. The light grey marked region marks the results that refer to the "main" date, which can be selected by simple clicking on a date in the timeline above (see 6).By clicking with the mouse into a sub-window, a window opens and the information is shown larger scaled. In this window, the user can modify the size/scalefactor of the value axis and also the time axis.
The main window provides the most detailed view on the data. First, its being used to select a patient by his/hers name (and some additional information) from a list returned by the search process. The main purpose of this window is to display detailed information of the test results. The user of this application drags and drops a test (or tests) he is interested in into this field to display it. This field has three sections aligned vertically filling up its full height and aligns the tests as follows. The upper and lower sections are being separated from the middle section by dashed lines. Additionally, they are filled with a gradient color. The upper sections holds test with the result with a value that lies above the top limit of the normal interval for this test. The center part holds all the tests with a result that lies in the normal interval. The last bottom part displays all tests that lie underneath the minimum of the normal interval. Each test is being displayed as a circle with a certain background color with some textual information. This information holds the name of the test, the value of the test and the units for this value. This field provides only information about the test results obtained during one test. The user can see these results in a time context in the window beneath this one (marked as 6 in the picture above). He can also navigate through multiple tests by using this window.  
<br><br>
=== ad 5: Medical Notes ===
<b>ad 5: Medical Notes </b><br>
Martha
Each circle in the main window, which refers to a single result of an examination, can be selected by mouseclick. A texteditor opens, and the user can write some important notes for a specific examination result (see Figure 2 + 3). By clicking on the name of the doctor in the notices window, some informations about the doctor, as hospital, department etc. are shown in the same window.
=== ad 6: Timeline / Graph ===
<br><br>
This window displays time series lines that reference the tests being picked by dragging a test from the left window into the main window. Its being used for two purposes. The first is to put the test results into a time context. The user can follow the progress of the selected tests, search for anomalies or/and find correlations. The other purpose of this window is to navigate through the time using it as time line. The user needs to select a certain date (dates in this window reference test dates) which makes the application reload the data in the main window. The color of each line in this graph is the same as the background color of this test in the main window. The user can also enlarge the region of this window to get a better representation of the data. The leftmost and rightmost arrows are being used to scroll through the the graph.
<b>ad 6: Timeline</b><br>
=== ad 7: Menu ===
As stated before, the main window displays examination results over time. On the timeline, the user can select a specific date by simply clicking on it with the mouse. The selected date is the date of interest. The font of the selected date appears a bit larger than the fonts of the dates beside. Values that were obtained at this date, are shown in the grey marked area of the main window. The user is also able to scroll through the time by using a scroll bar (the arrows positioned on the left and right of the timeline). By scrolling, results of the examinations in the main window get adjusted - according to the time interval that is currently shown int the timeline. Below the timeline are two arrows, that point to the left and right. The user can scale the shown timeline interval by pulling the arrows to the left or to the right (with the mouse). In this manner, the user can see and compare more values obtained at distinct dates (see Figure 2).
<br><br>
<b>ad 7: Menu </b><br>
By pressing the "Menu"-Button, a pop-up window opens and some options are listed. <br>
By pressing the "Menu"-Button, a pop-up window opens and some options are listed. <br>
The options:
The options:


* <b>Print shown diagnostic results:</b> The shown section in (4) and a table which contains informations for each examination can be printed.
* <b>Print shown diagnostic results:</b> The shown image the main window and a table which contains informations for each examination can be printed.
* <b>Print line chart:</b> The shown section in (5) can be printed. Additionally, the user has the option to define the from/to-dates new. A list of the shown examinations and related information is printed too (in form of a table).
* <b>Print whole diagnostic report:</b> The whole set of diagnostic findings of the picked date (which is displayed in the main window)can be printed in form of a table. In each row, one examination, it's result, the unit of the result and the range of optimal result values is put column by column. The user has also the possibility to choose another date for printing the related diagnostic findings.  
* <b>Print whole diagnostic report:</b> The whole set of diagnostic findings (not only the shown examinations in 4) of the picked date (in 4) can be printed in form of a table. In each row, one examination, it's result, the unit of the result and the range of optimal result values is put column by column. The user has also the possibility to choose another date for printing the related diagnostic findings.  
* <b>Send Message/Data:</b> The user can share notes or views with stuff of the same or other hospitals.
* <b>Options:</b> The user can customize specific functions of the program (e.g. personal layout etc.).
* <b>Help:</b> The user can read the documentation of the software and instructions of software usage.
<br><br>


<br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/><br/>
XXX '''REMOVE FROM HERE''' XXX
== Aufgabenstellung ==
Aufgabe ist das Design einer interaktiven Visualisierungsapplikation zur Darstellung und Exploration <br>(des zeitlichen Verlaufs) von Laborwerten einer Blutuntersuchung. BenutzerInnen, Einsatzzweck, Tasks, etc. <br>sollen von Euch selbst festgelegt und beschrieben werden.
[http://ieg.ifs.tuwien.ac.at/~aigner/teaching/infovis_ue/infovis_ue_aufgabe4.html Beschreibung der Aufgabe 4]
=== Beispiele für derartige Datensätze ===
[[Image:20061115_blutuntersuchung.jpg]]
[[Image:20070920_blutuntersuchung.jpg]]
XXX '''REMOVE TO HERE '''XXX
== Links ==
== Links ==


* [[Teaching:TUW_-_UE_InfoVis_WS_2007/08|InfoVis:Wiki UE Homepage]]
* [[Teaching:TUW_-_UE_InfoVis_WS_2007/08|InfoVis:Wiki UE Homepage]]
Line 80: Line 71:


*[[Teaching:TUW - UE InfoVis WS 2007/08 - Gruppe 07|Gruppe 07]]
*[[Teaching:TUW - UE InfoVis WS 2007/08 - Gruppe 07|Gruppe 07]]
==References==
*[Few, 2004a] Elegance Through Simplicity. Retrieved at: December 09, 2007. http://www.intelligententerprise.com/showArticle.jhtml
*[Few, 2004b] Show Me the Numbers: Designing Tables and Graphs to Enlighten, Analytics Press, 2004, Chapter 7 - General Design for Communication.
*[Tufte et al., 1999] Tufte Design Principle, Retrieved at: December 09, 2007. http://ldt.stanford.edu/ldt1999/Students/mizuno/Portfolio/Work/reports/tufte/ed229c-tufte-outline.html

Latest revision as of 16:22, 29 February 2008

Assignment description[edit]

Description of the fourth task

Assigned data / model[edit]

Description of the application field[edit]

The two tables above show medical data, which are two diagnostic findings of a patient recorded at different dates. In each table row, one examination, it's result, the unit of the result and the interval of "optimal" values are shown. The examination results descripe the amount of chemical compounds or substances in the patient's body. Diagnostic findings play an important role for medical diagnosis.
The first table consists of more rows than the second. Perhaps less substances in body were examined or the tableview shows only a short section of the results.

Each table has four columns. The names of the examinations and the name of the substances the patient has been tested on, respectively, are written in the first column. The result values - numerical values - for each examination are denoted in the second column. Some values are floating point numbers, others are integers. The reason of using floating numbers is, that small variances of the values may cause high differences in diagnosis. One examination has more than one result ("Differential Blutb."), but we don't know the meaning of this results. Some of the cells in this column are marked with "*". This indicates that a value does not reside in an interval (specified in the fourth column). In the third column, the units of the results are shown. The fourth column lists the intervals of reference values - the "optimal" values - for each examination. If a value resides within an interval, the result is "good" - which means that the patient has no lack and no overrun of a specific element.
Multiple forms of intervals are being used, e.g:

  • bis 200
  • 220-480
  • <66

The data is sorted ascending using the first column (the name of the test). None of the values can be classified as nominal or ordinal. We are only dealing with floating point numbers (continuous) and integers (discrete). There is also no hierarchy in this data. It's actually a list of key value pairs where the value-part can contain one or multiple values and an information about the units used and reference.

But the use of diagnostic findings in practice is slightly different. They are not being used as standalone tables with some test results. The values are being examined how they change in time. This fact changes the dimensionality of all data (test key-value-pairs) from 1-dimensional to 2-dimensional because they usually change over time. This is the main issue why we would suggest a different presentation of this data in a more clearer, simpler and legible way.

Target group[edit]

The target group (the data has been produced for) is definitely hospital/medical staff. We think that people who do not have any medical knowledge, are not able to make a diagnosis based on the data. They can basically determine which test results do not reside within the reference intervals. The data is very private and should only be seen from people authorized people - like medical stuff. Of course the application can also be used to give the patient an overview of his diagnosis data. For that purpose it is important that the attending physician is at present and can present the visualised data with several views and funded knowledge. For that reason we try to display the given data as intuitive as possible to ease the handling of the appication for the medical stuff aswell as patients that are assisted by a physican should be able to understand basic coherences and can therefore gain a better understanding for ongoing inquiries.

Concept & Mockup for: The DIAMONT application[edit]

We named our application DIAMONT which stands for DIAgnosis MONitoring Tool. This tool has been designed for medical staff, i.e. doctors and nurses. It shall provide a more easier legible and better organized view on the data from the test sets. The application is made as a desktop application or as a PDA (handheld-gadget) application. The idea is, to share the data with all hospitals and medical institutions. Therefore, the software has to be installed in each medical institution and the data itself is located on a remote server.

In the following, we will provide a detailed description of DIAMONT's interface. The red colored numbers mark the parts of the interface we want to describe in detail.


Figure 1


Figure 2: point marked with black circle can be edited in the medical notes sheet (see Figure 3)


Figure 3

ad 1: Patient Search
The patient search section allows to search for specific patients. By simply typing in a name (or other inforamtion which can be used to identify a patient) and pressing the "Suche Patienten"-Button, the application displays a list of records in the main window. By single-clicking a row, more detailed information about the selected patient is shown in the main window, which is for this purpose an "Patient view" window. Data of the patient can be: name, date of birth, dates of hospitals stays etc. Double-clicking a row will load the patients test values to the list of available tests (see 3).

ad 2: Patient information
This section displays the patient's name and birth-date. More detailed information about the patient is given (in the main-window) when the name of the patient is clicked with the mouse. The information is shown then in the main window and "Patiens view", respectively. There, the patient's data can be modified.

ad 3: List of available tests
The list of available tests shows the names of the tests/examinations that are represented in the dataset. By dragging a test into the main window (see 4), the examination information is visualised in the main window. Values that reside below or above the optimum range - at the chosen date - (see 6) are highlighted red. Values that are within the reference range are black. This feature allows the user to quickly scan the tests for outstanding values, that might be interesting for further research.

ad 4: Main Window
The user can drag and drop one or more tests (the amount is not limited) from the list (see 3) into the main window and the sub-windows, respectively. Empty sub-windows (where no examinations are dropped)show a message that says, that an examination can be dragged and dropped into this window. Each sub-window, which contains an examination, displays detailed information of the tests/examinations over time. The values obtained over time are visualised with a graph (line chart). The sub-windows are divided by horizontal lines into three parts. Examination values that reside above the reference interval are drawn at the top. The part in the middle shows values that reside within the interval. The bottom part shows test results that reside below the reference interval. Additionally, the values are marked with colored circles: red, if the value is too low or too high; green, if the value resides within the optimal interval. The light grey marked region marks the results that refer to the "main" date, which can be selected by simple clicking on a date in the timeline above (see 6).By clicking with the mouse into a sub-window, a window opens and the information is shown larger scaled. In this window, the user can modify the size/scalefactor of the value axis and also the time axis.

ad 5: Medical Notes
Each circle in the main window, which refers to a single result of an examination, can be selected by mouseclick. A texteditor opens, and the user can write some important notes for a specific examination result (see Figure 2 + 3). By clicking on the name of the doctor in the notices window, some informations about the doctor, as hospital, department etc. are shown in the same window.

ad 6: Timeline
As stated before, the main window displays examination results over time. On the timeline, the user can select a specific date by simply clicking on it with the mouse. The selected date is the date of interest. The font of the selected date appears a bit larger than the fonts of the dates beside. Values that were obtained at this date, are shown in the grey marked area of the main window. The user is also able to scroll through the time by using a scroll bar (the arrows positioned on the left and right of the timeline). By scrolling, results of the examinations in the main window get adjusted - according to the time interval that is currently shown int the timeline. Below the timeline are two arrows, that point to the left and right. The user can scale the shown timeline interval by pulling the arrows to the left or to the right (with the mouse). In this manner, the user can see and compare more values obtained at distinct dates (see Figure 2).

ad 7: Menu
By pressing the "Menu"-Button, a pop-up window opens and some options are listed.
The options:

  • Print shown diagnostic results: The shown image the main window and a table which contains informations for each examination can be printed.
  • Print whole diagnostic report: The whole set of diagnostic findings of the picked date (which is displayed in the main window)can be printed in form of a table. In each row, one examination, it's result, the unit of the result and the range of optimal result values is put column by column. The user has also the possibility to choose another date for printing the related diagnostic findings.
  • Send Message/Data: The user can share notes or views with stuff of the same or other hospitals.
  • Options: The user can customize specific functions of the program (e.g. personal layout etc.).
  • Help: The user can read the documentation of the software and instructions of software usage.



Links[edit]

References[edit]