Despite the remarkable advancements in artificial intelligence (AI), studies have found that it may not be able to improve the accuracy of medical diagnoses. It is therefore vital that next generation computer-aided diagnosis algorithms need to be both interactive and highly accurate in order to utilize the true potential of AI in improving medical diagnosis.

The University of Houston (Houston, TX, USA) has recently been awarded a grant from the National Cancer Institute for their upcoming project of creating a new AI system that will focus on improving diagnostics for lung cancer. This project plans on developing an AI-human collaboration framework, which will utilize eye-gaze tracking, intention reverse-engineering and reinforcement learning to determine when and how an AI system should interact with radiologists in order to make a medical diagnosis.

The primary focus of this project is to create a user-friendly and minimally interfering interface which will enable radiologist-AI interaction. It will be focusing on two major clinical applications: detection of lung nodules and pulmonary embolism. Lung cancer ranks as the second most common cancer, and pulmonary embolism is the third most common cause of cardiovascular death. This project will further investigate questions that have been largely under-explored, such as when and how AI systems should interact with radiologists and how to model radiologist visual scanning processes.

“Studying how AI can help radiologists reduce these diseases' diagnostic errors will have significant clinical impacts,” said Hien Van Nguyen, University of Houston associate professor of electrical and computer engineering, who is leading the project. “Our approaches are creative and original because they represent a substantive departure from the existing algorithms. Instead of continuously providing AI predictions, our system uses a gaze-assisted reinforcement learning agent to determine the optimal time and type of information to present to radiologists. Our project will advance the strategies for designing user interfaces for doctor-AI interaction by combining gaze-sensing and novel AI methodologies.” 

http://m.allmycbs.com/en/index.asp