The accuracy of an automasking algorithm in plantar pressure measurements

Masking algorithms provide a way to analyze plantar pressure parameters based on distinct anatomical regions of the foot. No study has addressed their accuracy. The purpose of this study was to determine the accuracy of the Novel® ten-region standard masking algorithm in both dynamic and static measurements in normal feet. Static and dynamic plantar pressure measurements were collected from ten normal subjects (20 ft) with and without 10-mm radiopaque markers placed under the first through fifth metatarsal heads, fifth metatarsal base, and first proximal phalanx. The automask was then applied to subdivide the foot into distinct anatomical areas. Weight-bearing AP radiographs were obtained with and without markers. Plantar pressures and radiographs were overlaid. The percent accuracy of each marker within its appropriate mask region was calculated. The average accuracies of the automasking algorithm regions for dynamic and static measurements, respectively, were 98.8% and 90.4% (1MH), 89.9% and 80.6% (2MH), 98.6% and 81.4% (3MH), 96.8% and 82.3% (4MH), 93.1% and 80.8% (5MH), 97.3% and 92.5% (5MB), and 91.2% and 64.2% (1PPH). Marker presence did not alter foot structure or function as determined by intermetatarsal angles (range, p = 0.361 to p = 0.649) and the center of pressure excursion index (p = 0.727), respectively. The automasking algorithm accurately identifies most foot regions in normal feet, particularly in gait. Such accuracy may be reduced in the setting of foot deformity. Understanding the accuracy of masking algorithms may help guide the interpretation of plantar pressure measurements and ultimately both conservative and operative treatment decisions.