Report Title: Biological Imaging Informatics and Clinical Pathology Diagnosis–the Development in the Digital and AI Era
Date: January 2, 2024 (Tuesday), 15:00-16:30
Location:Zhenzhen Building Room 404, Smart Classroom
Speaker: WeiMiao Yu
Personal Introduction:
Prof. YU Weimiao (Principal Scientist), A*STAR
Head of Computational Digital Pathology Lab
Bioinformatics Institute (BII)
Head of Computational and Molecular Pathology Lab
Institute of Molecular and Cell Biology (IMCB)
Agency of Science Technology and Research (A*STAR)
Adjunct Professor
School Artificial Intelligence,
Nanjing University of Information Science and Technology (NUIST)
Adjunct Professor
School Biological Science (SBS),
Nanyang Technological University
Abstract
In recent years, the advancements in image-based deep learning technologies have significantly propelled the field of Artificial Intelligence (AI)-based pathological diagnosis. Numerous promising research outcomes have emerged. However, the translation of these advancements into real-world clinical applications remains limited. The primary hindrance lies in an array of challenges and bottlenecks, encompassing both technical and non-technical aspects. The field still lacks certain critical components required for comprehensive development. In this sharing session, we will conduct a thorough review of the field's current status while scrutinizing the obstacles, challenges, and bottlenecks impeding progress. Additionally, we will delve into understanding the ecosystem that underpins the industry's growth. Three pivotal concepts, optimization, generalization andvariation,serve as linchpins for successfully deploying AI models in clinical settings. I will share fundamental principles that can guide us in achieving optimization and generalization when designing our studies. Furthermore, we will explore five essential and practical solutions aimed at expediting the development of clinically applicable AI models in pathological diagnosis. This will set the stage for the accelerated growth of AI-based pathological diagnosis in the years to come.
Bio
Dr. Yu Weimiao obtained his Ph.D. from National University of Singapore (NUS) in 2007, majoring in image processing and machine vision. He joined Agency of Science, Technology and Research(A*STAR) in 2007 and moved his research position to Institute of Molecular and Cell Biology (IMCB) in 2010. He is currently heading Computational Digital Pathology Lab (CDPL) in Bioinformatics Institute (BII) and Computational Molecular & Pathology Lab (CMPL) in IMCB to deepen and extend the R&D with clinical and industrial partners. His research interests are Computational Biomedical Image Analysis and Quantitative Imaging Informatics in biomedical image. Dr. YU’s research outcomes were published in top international peer reviewed journals, such as Nature Cell Biology, Nature Communication, Current Biology, etc. His previous research outcomes were successfully commercialized and made an impact in the diagnosis and drug discovery. To enhance the application of machine learning and AI in clinical diagnosis/prognosis, he founded a biotech company in Singapore, known A!maginostic Pte. Ltd. He established a world class platform for the immunodiagnosis at tissue level. The platform allows the researchers, clinicians and pharma to profile the patient immune signature for the purpose of diagnosis, prognosis and drug response study.
