Automated pressure map segmentation for quantifying phalangeal kinetics during cylindrical gripping

Inverse dynamics models used to investigate musculoskeletal disorders associated with handle gripping require accurate phalangeal kinetics. Cylindrical handles wrapped with pressure film grids have been used in studies of gripping kinetics. We present a method fusing six degree-of-freedom hand kinematics and a kinematic calibration of a cylinder-wrapped pressure film. Phalanges are modeled as conic frusta and projected onto the pressure grid, automatically segmenting the pressure map into regions of interest (ROIs). To demonstrate the method, segmented pressure maps are presented from two subjects with substantially different hand length and body mass, gripping cylinders 50 and 70 mm in diameter. For each ROI, surface-normal force vectors were summed to create a reaction force vector and center of pressure location. Phalangeal force magnitudes for a data sample were similar to that reported in previous studies. To evaluate our method, a surrogate was designed for each handle such that when modeled as a phalanx it would generate a ROI around the cells under its supports; the classification F-score was above 0.95 for both handles. Both the human subject results and the surrogate evaluation suggest that the approach can be used to automatically segment the pressure map for quantifying phalangeal kinetics of the fingers during cylindrical gripping.