AI Quantum Entangled Pet
Project Overview
PET imaging is the primary diagnosis tool for Cancer and Alzheimer’s. It has global market value~$500Bn. York researchers led a recent paradigm shift in our understanding of the two gamma photons detected in PET, they have quantum entanglement (QE) with each other, and it is maintained even when the photons scatter in the patient (90% of scan events).
The York group is in a unique position to model and extract information from these discarded events for the first time. The scattered photon simulation data will train an AI model - to determine both the annihilation and scatter site maps.
What were the key results of your research project?
- Created two models to predict scatter and annihilation maps.
- Created a function to obtain Lines of Response(LORS) from the prediction and another function to perform attenuation correction on the LORS.
- Created a function to create an image using back-projection out of the LORS. -Created a proof of concept for AI-corrected PET.
How do you feel you have benefited from completing this internship, and has it made you consider future career paths?
have further developed my independent reteach skill as well as my programming skill, especially using ML libraries such as PyTorch. Also developed communication skills working with other teams to reach deadlines.
I have found it rewarding to apply computational methods to physical systems. The process of developing code has also been fun, therefore, I would definitely consider a future career as a research software engineer in the field of physics.
Download presentation slides