Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science


Computer Science


Charles Ling


In the case of breast cancer, according to the Nottingham Grading System, counting mitotic cells is an important indicator of tumour diagnosis and grading. Pathologists usually manually count mitosis from histopathology images to determine the cancer grade. This is a challenging and time-consuming procedure. In most recent works, different deep neural networks have been designed to detect the suspicious cells initially and count the number of them afterwards. However, these detection approaches have certain limitations including complicated structures, the detection performance is still not satisfactory, and the need of a large number of labeled images to train a satisfied model. In this paper, we modify and improve a popular one-stage object-detection deep network to facilitate the mitotic cells detection task. Our novel improvements include using different loss functions for cells of different sizes, designing new data augmentation methods, generating prior anchor boxes with approximate sizes by using an improved clustering algorithm, and so on. We validate our deep learning model on two public benchmark datasets named Mitosis Detection in Breast Cancer Histological Images (MITOSIS). The experimental results indicate that our method achieves the competitive results on MITOSIS-2012 dataset and on the MITOSIS-2014 dataset with faster inference speed. More importantly, we design an interactive system with a "correction and relearning" pipeline so that our system can relearn from a small number of slides from a new lab and achieve satisfactory results. We design a web portal (http://ai4path.ca/#/) where this online pipeline can be easily utilized by pathologists in Western Hospital Pathology Group(WHPG) and hopefully in the future, by all pathologists in the world.

Summary for Lay Audience

Mitotic cells detection is an important step in the pathology domain. It is mainly used to diagnose and prognose cancer in different regions of our body, for instance, breast, glioma, and melanoma etc. However, most of these examinations by pathologists under microscopes are still manual work which is time-consuming, challenging and subjective. Therefore, it is significant to provide a real-time computational examination tool for pathologists. In recent years, with the development of Convolutional Neural Network (CNN) in the computer vision domain, computers can have the ability to detect objects in images. The success in the field of computer vision has attracted the attention of researchers in the pathology domain. Previous attempts are developed based on constructing manual features and obtaining features from deep neural networks. Both of these approaches are widely deployed to accomplish the mitotic cells detection task and achieve certain results. However, the detection performance of these current methods are still not satisfactory.

In this thesis, we primarily deal with the mitotic cells detection task through modifying and improving a popular object detection method. In addition, we extract the learned knowledge from previous work as prior information. We integrate the prior information in our model to create a robust real-time tool for pathologists to examine the detected results.