Printhotics - Deep learning to optimize and personalize 3D printed pediatric medical devices
BACKGROUND: Ankle-foot orthoses (AFOs) are a highly effective non-surgical treatment for children with walking difficulties in many conditions, such as cerebral palsy (1 in 500 births), muscular dystrophy (1 in 3000 births) and peripheral neuropathy (1 in 2500 births). Traditionally, AFOs are handmade by vacuum forming polypropylene plastic over a plaster of Paris cast, relying on manual labor with unacceptably long lead times.
GAP: To bring these devices to growing children faster, we are exploring the use of 3D scanning and 3D printing to produce AFOs that function more effectively and that are designed to improve compliance. This project aims to digitize the craftsmanship of manual plaster modification involved in the production of AFOs by using 3D scanning and machine learning.
HYPOTHESIS: I hypothesize that the plaster additions made during the AFO fabrication process are predictable and occur primarily at key regions of the foot and ankle.
METHODS: Plaster casts will be 3D scanned and fed into a predictive algorithm. We will include the casts of any child presenting to the Children’s Hospital at Westmead for AFOs.
IMPACT: The results from this study will close the loop between 3D scanning patients and 3D printing AFOs, enabling us to bring these critical medical devices to the children that need them most.