- About Us
- Welcome Message
- Vision
- Brief History
- Staff
- Contact Us
- Faculty / Research
- Faculty
- Research Labs
- Academics
- Curriculum
- Course Guide
- Graduate School
- Graduation Project
- Admissions
- Admissions
- Academic Affairs
- Academic Calendar
- Scholarships
- Graduation Requirements
- Student Clubs
- Academic Information
- Facility Reservations
- Community / News
- NEWS
- Notices
- Event
- Job Openings
- Forms
- KU Q&A
Community / News
NEWS
Prof. Jin Tae Kwak’s Research Team at Korea University Develops AI for Gastric Pathology Diagnosis Using Sequential Question Answering That Thinks and Asks Like a Doctor
관리자 2026.08.12 Views 18
▲ Professor Jin Tae Kwak, Korea University
A research team led by Professor Jin Tae Kwak of the Korea University School of Electrical Engineering has developed AUGUST, an artificial intelligence model that diagnoses diseases by emulating the sequential reasoning process of practicing pathologists. The study was recently published in npj Digital Medicine (https://www.nature.com/articles/s41746-026-03107-1), a leading international journal in healthcare science, earning global recognition for the achievement.
Conventional digital pathology AI systems often treat diagnostic tasks independently and fail to account for the strict hierarchical dependencies in gastrointestinal pathology, where higher-level diagnostic decisions determine subsequent, more specific diagnoses. As a result, their conclusions may lack logical coherence. To overcome this limitation, the research team introduced a “sequential question-answering framework,” in which the AI repeatedly generates tailored questions for each successive step based on its preceding diagnostic findings. This approach enables the model to analyze massive whole-slide images (WSIs) in context and arrive at a final diagnosis through a clinically valid reasoning pathway-much like a human physician.
The study holds significant academic value because it addresses the reliability limitations of conventional black-box AI by transparently presenting each step of the diagnostic reasoning process. Validation using a large-scale dataset of more than 23,000 WSIs demonstrated that AUGUST substantially outperformed existing state-of-the-art AI models in both accuracy and consistency.
The research team expects AUGUST to serve as a valuable clinical decision support tool that improves diagnostic accuracy and reliability by transparently presenting not only its final diagnosis but also the reasoning behind it. The study was conducted through close interdisciplinary collaboration among medical and AI experts, including researchers from The Catholic University of Korea Uijeongbu St. Mary’s Hospital.
Conventional digital pathology AI systems often treat diagnostic tasks independently and fail to account for the strict hierarchical dependencies in gastrointestinal pathology, where higher-level diagnostic decisions determine subsequent, more specific diagnoses. As a result, their conclusions may lack logical coherence. To overcome this limitation, the research team introduced a “sequential question-answering framework,” in which the AI repeatedly generates tailored questions for each successive step based on its preceding diagnostic findings. This approach enables the model to analyze massive whole-slide images (WSIs) in context and arrive at a final diagnosis through a clinically valid reasoning pathway-much like a human physician.
The study holds significant academic value because it addresses the reliability limitations of conventional black-box AI by transparently presenting each step of the diagnostic reasoning process. Validation using a large-scale dataset of more than 23,000 WSIs demonstrated that AUGUST substantially outperformed existing state-of-the-art AI models in both accuracy and consistency.
The research team expects AUGUST to serve as a valuable clinical decision support tool that improves diagnostic accuracy and reliability by transparently presenting not only its final diagnosis but also the reasoning behind it. The study was conducted through close interdisciplinary collaboration among medical and AI experts, including researchers from The Catholic University of Korea Uijeongbu St. Mary’s Hospital.
File
Engineering Building #412, College of Engineering, 145 Anam-ro, Seongbuk-gu, Seoul, 02841, KoreaTEL : 02-3290-3210, 3211, 3213, 3214, 4592FAX : 02-921-0544
Copyrights ⓒ 2021 KOREA University. All Rights Reserved.
- Family Site