Efficient symptom inquiring and diagnosis via adaptive alignment of reinforcement learning and classification

Hongyi Yuan,Sheng Yu

DOI: https://doi.org/10.1016/j.artmed.2023.102748

IF: 7.011

2024-01-07

Artificial Intelligence in Medicine

Abstract:Medical automatic diagnosis aims to organize real-world diagnostic processes similar to those from human doctors and to achieve accurate diagnoses by interacting with patients. The task is formulated as a sequential decision-making problem with a series of information inquiry steps (asking about symptoms and ordering examinations) and the final diagnosis. Recent research has studied incorporating reinforcement learning for information inquiry and classification techniques for disease diagnosis, respectively. However, studies on efficiently and effectively combining the two procedures are still lacking. To address this issue, we devised an adaptive mechanism to align reinforcement learning and classification methods using distribution entropy as the medium. Additionally, we created a new dataset for patient simulation to address the lack of large-scale evaluation benchmarks. The dataset is extracted from the MedlinePlus knowledge base and contains significantly more diseases and more comprehensive symptom and examination information than existing datasets. Experimental evaluation shows that our method outperforms three current state-of-the-art methods on different datasets by achieving higher medical diagnostic accuracy with fewer inquiring turns.

engineering, biomedical,computer science, artificial intelligence,medical informatics