AI Model Development for Automatic Extraction of Electronic Medical Record Information and Prediction of Diagnosis for Respiratory Diseases

Electronic medical records contain a wealth of valuable information, and by utilizing it in machine learning, it is believed that highly accurate AI models can be developed. However, the current medical records include a significant amount of unstructured or non-numeric data, which cannot be directly used for machine learning.
We are focusing on electronic medical records for respiratory diseases and developing a method to automatically convert the unstructured and non-numeric data in the records into structured numeric data that can be utilized for machine learning. Using this conversion method, we are planning to extract useful information from the electronic medical records of respiratory diseases and ultimately develop a model that can accurately predict the final diagnosis.

Development Section in Charge

AI Development for Medical Diagnosis and Treatment Section

Principal Investigator

KUDO Kohsuke, Professor, Department of Diagnostic Imaging, Hokkaido University Graduate School of Medicine

University-Affiliated Researchers

HIRATA Kenji, Associate Professor, Department of Diagnostic Imaging, Hokkaido University Graduate School of Medicine
TANG Minghui, Specially Appointed Assistant Professor, Department of Diagnostic Imaging, Hokkaido University Graduate School of Medicine
HAN Feng, PhD Student, Department of Diagnostic Imaging, Hokkaido University Graduate School of Medicine
ZHANG Ziheng, PhD Student, Department of Diagnostic Imaging, Hokkaido University Graduate School of Medicine
OGASAWARA Katsuhiko, Professor, Department of Health Sciences and Technology /Digital Health Innovation, Faculty of Health Sciences

Non-University-Affiliated Researcher

YAGAHARA Ayako, Associate Professor, Hokkaido University of Science Faculty of Health Sciences Department of Radiological Technology