A New Framework for denoising MRI scans; Access to Medical Imaging in Rural India 7 views

Though it can be crucial to diagnosis, medical imaging is expensive. How can medical imaging be made more accessible? How can MRI image quality be enhanced without compromising accuracy? Solving this problem would improve MRI accessibility by making substandard scanners and resulting scans usable. No established software-based solution exists. Prototype testing was used in which an energy-based model (EBM) was trained on a dataset consisting of brain slices with the objective to reconstruct damaged slices and generate new slices. General improvements in speed and accuracy were observed for slice generation and favorable parameter configurations were identified. Diffusion models such as EBMs show promise for problems of this nature. In rural India, how are physicians, patients, advocacies, medtech companies, NGOs, and policymakers influencing patient access to medical technology? Since most Indians reside in rural regions, improving this infrastructure would expand access to medtech nationwide. As for-profit enterprises, medtech companies generally favor profitable but expensive state-of-the-art equipment, including imaging systems, and ambitious physicians are drawn to the well-equipped urban hospitals that have them. Rural health centers, however, must economize, often by foregoing state-of-the-art medtech. While policymakers, advocacies and NGOs have a compensatory effect, their resources are too limited to equip rural hospitals with state-of-the-art equipment, and to improve care, most favor reaching more people with basic healthcare over reaching fewer people with the best equipment.

BS (Bachelor of Science)

Computer Science; Machine Learning; MRI

School of Engineering and Applied Science Bachelor of Science in Computer Science Technical Advisor: Mathews Jacob STS Advisor: Peter Norton

All rights reserved by the author (no additional license for public reuse)

2025-12-11