AEHRC’s Aaron Nicolson takes out top prize in international imaging competition

May 27th, 2024

The AI model Dr Nicolson developed gained him first prize in the competition run by the Association for Computational Linguistics.

The competition involved developing an AI model to produce the most optimal radiology reports within a limited amount of time. 

Aaron is working on developing models for automated chest x-ray report generation. 

In explaining the potential impact of his work Dr Nicolson said, “AI has the potential to improve health services, and in particular better support health professionals by assisting them in their clinical workflows.” 

“Automated report generation for chest X-rays could reduce radiologist burnout and create space for them to provide more robust patient care. This research is a step in the right direction towards an assistant for radiology.” 

There were four tasks in total, and Aaron beat the other nine international competitors in 3 of the 4 tasks.  

Another AEHRC scientist, Dr Jinghui Liu, also came third in one of the tasks.  

The multimodal language model that placed first in the radiology report generation task.