ORTHOSOLID
AI to digitally develop prosthetic sockets based on a 3D scan
What
People who undergo an amputation of their lower leg can use a prosthesis to help them walk again. A kind of silicone sleeve (a liner) is pulled around their stump, a prosthetic socket is placed around it and the prosthesis itself is attached to that socket. The prosthetic socket (of the lower leg) must be custom made.
Currently, prosthetists do this on the basis of a physical, plaster copy of their patients' stump and that copy is then corrected. This is done by applying and removing material from specific places on the plaster copy. The prosthetist uses the modified plaster model to create the prosthetic socket.
In this project, Mobilab & Care designs an alternative way to make such a customized prosthetic socket, namely using 3D scan and artificial intelligence (AI).
Goal
The Orthobroker company is now developing software that allows prosthetists to scan the stump of their patients and, based on the 3D scan, create a digital design of the final prosthetic socket. That model can then be printed.
To make this software user-friendly, a predictive model will be developed using artificial intelligence (AI) that will suggest corrections to the prosthetist after scanning the residual limb.
Relevance
The last few years, the use of 3D scanning and 3D printing in the production of orthopedic aids has become increasingly useful. This method allows the prosthetist to work faster, produce less waste, and create multiple identical models.
However, it is not easier for a prosthetist to suddenly switch from the manual method he/she has been using for years to a digital one. That’s why artificial intelligence is used in this project. By using such a method, the prosthetist receives a proposal of corrections that must be carried out digitally. These can then be further adjusted to the patient’s specific needs.
Our role
The role of Mobilab & Care is to develop the algorithm that predicts the corrections to be made on the 3D scan of a stump.
The algorithm learns which corrections need to be made based on examples of corrections made by prosthetists on plaster casts of stumps. To do this, M&C researchers go to various orthopedic companies and different prosthetists to make a scan of both the patients’ stumps and the corrected plaster model during their daily clinical practice. Characteristics are derived from these, which, together with other patient data, are used to create the predictive model. This predictive model will be used in Orthobroker software. This allows prosthetists to intuitively create a digital design of the final prosthetic socket based on the 3D scan of their patients’ stumps.
Researchers
Kris Cuppens
Draws energy from processing, analysing and presenting data. Always in the mood for a board game.
Veerle Creylman
Loves to chat about orthoses, prosthetics and how to make them as optimal as possible, for every patient!
Inge Van den Herrewegen
Forever on the path to essence and truth. Inspires at the bridge between theory and practice, science and connection.
Hajar Ghaem
Applied Data Scientist with focus on AI and machine learning technology in the health domain.
Partners
