Three-dimensional differences in plantar surface shape captured by methods used for custom accommodative insole design.

Background: The patient-specific shape of custom accommodative insoles for individuals with diabetes provides full foot-to-insole contact, offloading areas with high plantar pressures and reducing ulceration risk. To design the insole surface, plantar surface shape is captured, traditionally with a foam crush box impression or more recently with 3D scans of the foot. Beyond discrete measurements of the foot, the overall plantar surface shapes obtained from these different methods have yet to be compared, however, differences in the shapes captured by these methods may affect the insole's surface geometry design and subsequent performance.

Methods: Plantar surface shapes of 12 individuals with diabetes were captured using a foam crush box, flatbed 3D foot scanner, and handheld 3D scanner. Foot length, width, arch height, and arch volume were measured from each shape-capture method and compared. Mesh-to-mesh distances between the foam crush box mesh and the direct scanning method meshes for each subject were calculated.

Results: Foot length and width measured from the foam crush box scan were greater than the foot length measured from the flatbed scan and handheld scan. The flatbed scan also measured a length and width greater than the handheld scan. Arch heights and volumes from the flatbed scan were less than the heights calculated from the foam crush box and handheld scan. Mesh-to-mesh distances for the flatbed scan and areas of the foot not in contact with the scanner were inferior to the corresponding areas in the foam crush box impression. For the handheld scan, the lateral hindfoot and midfoot were superior, and the medial forefoot was inferior to the foam crush box impression.

Conclusions: Different clinical methods used to capture foot shapes for the design of accommodative insoles may result in different plantar surface shape outputs and therefore impact custom accommodative insole design.