Skip to content

Development of machine learning methods for automatic generation of personalized 3D medical models and pre-operative planning and their experimental application by extending the functionality of the 3D Web platforms MICE and Smart3Dworks

 

Project no.: 02-020-K-0068

Project description:

Millions of orthopedic, neurosurgical, and maxillofacial surgeries are performed globally each year, with implants being utilized in the vast majority of these procedures. While standard implants currently dominate the market (used in 97-98% of cases), the share of personalized implants is gradually increasing and is statistically projected to reach 10% of all surgical operations in the future. A critical trend mandates that the center of gravity of surgical treatment shift toward comprehensive pre-operative planning, moving away from relying on decisions made within the operating room. This transition would fundamentally improve treatment quality and optimize costs by shortening operation times, reducing patient complications, and decreasing the need for revision operations. Unfortunately, many operations are still performed without a high-quality surgical plan, often because surgeons lack the time or competence to utilize accurate planning tools, sometimes resorting to inaccurate 2D instruments. Therefore, there is a strong need to improve pre-operative planning protocols to ensure surgeons can make informed decisions before the operation, rather than depending on implant suppliers or planning after the incision has been made. To address this need, the project plans to create an intelligent (automatic) system for designing personalized 3D implant and surgical guide models, based on machine learning. Additionally, the project will develop an intelligent system for generating and validating a detailed pre-operative plan utilizing machine learning, based on the patient’s personalized 3D anatomical data and established clinical reference points.

Project funding:

This research project is funded by the European Union Funds for the period 2021-2027 under the Measure No. 05-001-01-05-07 “Establishing a coherent system for the promotion of innovative activities” under the activity “Stimulating the supply of innovations” under the action “Investing in activities for the development of new high value added products and enabling researchers to participate in R&D activities of enterprises, promotion of intellectual property, early pilot production of new products being developed and preparation for the market” (region of Central and Western Lithuania)”


Project results:

The project plans to create: • An intelligent (automatic) system for designing personalized 3D implant and surgical guide models, based on machine learning. The system, based on the surgical plan and the personalized 3D anatomical model of the patient and the clinical reference points identified on the model, will automatically design the design and surface morphology of medical devices intended for the patient’s treatment. • An intelligent (automatic) system for generating and validating a pre-operative plan, based on machine learning, which, based on the personalized 3D anatomical data of the patient, the established clinical reference points characterizing both the patient’s anatomical features and his clinical pathology, and the 3D pre-operative planning performed by the doctor, will automatically prepare and document a detailed surgical treatment plan.

Period of project implementation: 2024-07-09 - 2027-04-30

Project coordinator: UAB "Ortho Baltic"

Project partners: Kaunas University of Technology

Head:
Darius Jegelevičius

Duration:
2024 - 2027

Department:
Biomedical Engineering Institute, Biomedical Imaging Laboratory