Meredith, in collaboration with Ulster University and axial3D has been shortlisted at the AI Ireland Awards ‘Best Application of AI in a Student Project’ – for her work on using AI to accelerate production of 3D printed cardiac models using machine learning, improving diagnosis and enabling surgeons to practice virtually before surgery.
It currently takes a trained medical visualization engineer at axial3D six hours to fully segment blood flow and cardiac tissue from cardiothoracic computerized tomography (CT) images. This causes bottlenecks within the 3D printing workflow, limiting the number of patients that could be helped by receiving a 3D printed model.
The purpose of this research, in collaboration with Ulster University and axial3D, and the model created was to automatically annotate blood flow within cardiothoracic computerized tomography images. By integrating machine learning into the company’s workflow, Meredith was able to successfully reduce the segmentation time; which is key for axial3D to reduce the amount of manual intervention.
This will positively impact surgeons by reducing the time taken in the turnaround of 3D printed anatomical models. This, in turn, aids in the diagnosis and treatment of patients, improving surgical planning and patient care.
By applying modern machine learning architectures it is possible to translate research models to an industrial setting for segmentation of cardiac blood flow in cardiac-vasculature images. Meredith has shown that machine learning can revolutionize medical 3D printing and support improved patient outcomes.
Gartner predicts that by 2021, 25% of surgeons will practice on 3D-printed models of the patient prior to surgery. As 3D printing becomes routine in surgical intervention, Meredith’s algorithm will aid in addressing this demand.
Join us in wishing Meredith the best of luck at the awards ceremony on November 20th in Dublin.