Innovative deep-learning based program for cervical cancer screening

The Data Science Institute is pleased to announce a research team led by Drs. Xiaoxiao Li (Electrical and Computer Engineering) and Gang Wang (Pathology and Laboratory Medicine) has been awarded the Postdoctoral Matching Fund from the DSI. This project aims to develop an end-to-end automatic deep learning-based cervical cancer screening pipeline that requires less labelling and addresses challenges in multi-institutional learning.

Summary: This project focuses on creating an end-to-end automated deep learning pipeline for cervical cancer screening that minimizes the need for extensive labeling and is generalizable across various medical centers. Cervical cancer ranks as the fourth most common cause of cancer-related deaths among women, claiming approximately 342,000 lives globally in 2020. In British Columbia, over 400,000 individuals undergo cervical cancer screening annually through Pap smear tests. Cytological image analysis is pivotal in the early detection and treatment of cervical cancer. However, current approaches are hindered by the need for extensive expert annotations, and their accuracy is compromised by the diversity in imaging techniques and the intricacies of the disease. Additionally, there is an absence of a privacy-centric mechanism to access data from multiple centers without sharing the data itself. Our research goals include exploring label-efficient solutions for cell segmentation, feature learning, and cytological image classification by employing cutting-edge self-supervised learning strategies and transfer learning from foundation models. Moreover, the project will investigate federated learning techniques to streamline the learning process and deployment while ensuring the adaptability of the developed method across different institutions. The successful completion of this project is expected to lower the medical diagnoses costs and facilitate the utilization of multi-center datasets for training, culminating in widespread benefits for healthcare systems throughout Canada.