Research Projects

India faces a significant challenge with diabetes, with 77 million adults affected and 25% of them at risk of developing diabetic foot disease. The diabetic foot disease is a serious complication and is being identified as the next public health crisis in India. Current therapeutic footwear lacks sufficient customization and optimized pressure distribution. In the current scenario, the emphasis is on diagnostic tools, therapeutic footwear, modular designs using conventional approaches. There is a pressing need for innovative, patient-specific solutions. The research proposes using Triply Periodic Minimal Surface (TPMS) structures for designing diabetic insoles. TPMS structures have tunable mechanical properties, and high energy absorption. It aims to demonstrate that the TPMS-based insoles, optimized through Finite Element Modeling (FEM) and AI, can provide an effective patient-specific solution to redistribute the plantar pressure that reduces ulcer risk and improves comfort. The research has a societal impact, significantly improving the patient well being and contributing significantly to the diabetic research. The TPMS geometries will be selected based on their implicit functions, CAD modeling precision, and ease of 3D printing. Appropriate material based on its mechanical properties and 3D printing compatibility. Using FEM, the mechanical properties of the TPMS structures, such as the compressive strength, energy absorption, fatigue strength, durability, etc. will be characterized under static and dynamic load to simulate activities like standing and walking. Parametric study will be performed to assess the effects of variables like unit cell size, relative volume, and wall thickness on the pressure distribution. Mechanical testing will be done to validate these models. The plantar pressure data of the patients (in collaboration with AIIMS Nagpur) will be collected across age groups, weights, and stage of the disease using pressure mats and pressure sensors. This data and the FE model developed earlier will guide the customization of the TPMS-based insoles for optimal pressure redistribution. Clustering algorithms will be used to categorize the plantar pressure data and identify any trends and correlations. This will help to create base designs for the insoles. CNN will be used to process the images of the pressure distribution which will help to tailor a patient-specific solution. A Recurrent Neural Network will be used to process the time series data of plantar pressure to predict the progression and location of ulcer. The research will produce a library of CAD compatible TPMS geometries with material recommendation and best 3D printing practices. A high-fidelity FE model, validated with mechanical testing, will allow faster prototyping and design optimization. This collaborative work between engineering and medical science has the potential to improve patient quality life and will contribute significantly in the diabetic research.