Technical Description of Computer Analysis of Pain Diagrams system (CAPD) used in Pain Survey Clinical Study (For Medical Professionals)

Pain shapes drawn by study subjects will be analyzed by computer to provide automated identification of body location and other statistical measures. These data will be analyzed together with body-location-specific data on the nature of the symptoms and the functional disability associated with pain. Subjects will be allocated to data-driven pain patterns of potential diagnostic value.

Key features of the computer analysis include:

• A novel TMT computer analysis system (patent pending) processes pain drawings provided by subjects. Subjects draw an outline of their pains in black on a cyan-colored, digitized, bilaterally symmetrical, coregistered human body diagram. They also mark the spot within each pain outline where pain is worst.

• Pain drawings provided by subjects are digitally captured using vector graphics

• Each pain diagram is computer-analyzed to provide measures of size (calculated pain area, Total Pain Area Load - TPAL, Local Pain Area Load - LPAL), centroid location, outline (perimeter), location point of worst pain, shape type, and anatomic areas occupied by pain. In addition, computationally simpler surrogates (such as application of an ellipsoid model to the pain shape with calculation of the long and short axes) may be used.

• Subject groups will then be compared using composite images of the group pain diagrams and statistical analysis of the derived variables.

The following figures show some of the results from computer analysis of pain drawings.

Figures

Patient Profile: Computer-generated figure of pain location and symptoms (Click thumbnail picture above to see larger image)

Lower Back Centroids and Meta-Centroid (Click thumbnail picture above to see larger image)

Lower Back Composite Image: Assumes Gaussian Distribution for Distances from Meta-centroid to Individual Pain Shape Edges for each Radian around the Meta-centroid  (Click thumbnail picture above to see larger image)

Patient-Recorded and Computer-Identified Shoulder Pain (Click thumbnail picture above to see larger image)

Pain Patterns: Shoulder, Hip, Knee, Back, Feet (Click thumbnail picture above to see larger image)

Poor Pain Shape Drawings: Computer Editing (Click thumbnail picture above to see larger image)

Centroid Masks: For Body-Area-Specific Location Assignment (Click thumbnail picture above to see larger image)

.