Study
PLCO
(Learn more about this study)
Project ID
PLCOI-1315
Initial CDAS Request Approval
Sep 5, 2023
Title
Histopathologic Image-based Deep Learning Classifier for Predicting Clinical Outcomes in Cancers
Summary
We have developed an artificial intelligence classifier using histopathologic imaging to predict progression-free survival in patients with high-grade serous ovarian cancer (HGSOC). Our approach involves two main steps: initially, a cell segmentation model is employed to identify cancer regions, followed by a multiple instance learning model to predict the aforementioned outcomes. In both internal and external validation sets (our in-house cohort and the TCGA cohort), our model successfully differentiated between favorable and poor prognosis groups (p < 0.0001). Model decisions were explained through visualization by pathologists and molecular analysis by bioinformatics, which bolstered reliability and helped further understanding of the biological behavior of each group. However, the number of patients in the evaluated cohorts are limited to generalize our results. With the help of PLCO datasets, we wish to increase generalizability and validate our prediction model for survival. In addition, we would like to further generalize our model to other cancer types showing high-grade histology.
Aims
1. To evaluate the performance of our survival prediction model in the PLCO cohort.
2. To identify histopathologic features associated with clinical outcomes or molecular characteristics, with the ultimate aim of creating predictive digital biomarkers using whole slide images in HGSOC.
3. To generalize our model developed for HGSOC to other cancer types.
Collaborators
Byungsoo Ahn, Department of Pathology, Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea
Damin Moon, Artificial Intelligence Research Center, JLK Inc., Seoul, South Korea
Chung Lee, Department of Pathology, Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea
Nam-Hoon Cho, Department of Pathology, Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea
Heung-Kook Choi, Artificial Intelligence Research Center, JLK Inc., Seoul, South Korea
Dongmin Kim, Artificial Intelligence Research Center, JLK Inc., Seoul, South Korea
