Classification for Pathological Images Using Machine Learning

An exploration in applying machine learning techniques in digital pathology

About the Project This is the final year project by Chi Ian Tang, a computer science student at the University of Hong Kong, under the supervision of Dr. S. M. Yiu.

In this project, we aim to explore the opportunities in applying machine learning techniques in medical context, especially in digital pathology. We developed a automated system to diagnose Bacterial Vaginosis infection, a common infection found in women at reproductive age, by examining the smears.

Methodology Overview

We approach the diagnosis problem with the following techniques:

Data Pre-processing

We start by calibrating the variations between microscopic images in the training dataset. This could include using Gaussian filters and granulometric analysis.

Image Segmentatioin

The normalised images are then segmented using differenet methods and a variety of blob detection algorithms including sliding window, DBSCAN and MSER, into interest areas which typically contains a patch of the image or a single bacterium.

Classification and Detection

In the first two phases of the project, a classifier which distinguishes between target bacteria types from the rest is trained. In the final phase of the project, a object detector is trained, predicting both the bounding boxes of the bacteria and the types of them. A number of different machine learning algorithms and models are used, including convolutional neural networks, residual networks and faster R-CNN.

Interpretation

The classified images are then collected, and the overall interpretation in the degree of infection is determined using the Nugent Score system.

User Interface

These are the user interface of the tools developed in this project.

Clicklable

Clicklable is a highly reusable data labelling tool for image classification and object detection purposes.

Users can choose different labelling modes including "Point" labelling or "Rectangle" labelling which facilitates both interest-area classification and object detection.

This tool is platform-independent, where any computer with Java™ Runtime Environment installed can run this tool with the most native interface.

Dignostic System for Bacterial Vaginosis

This diagnostic system is a all-in-one system for disgnosing Bacterial Vaginosis.

Users can load a trained machine learning model and starts the diagnosis by picking a smear image. Detailed results are drawn on the image and interpretations are also given. Users can also export the image for other purposes.

This tool is also platform-independent, where any computer with Python and the corresponding dependencies installed can run this tool with the most native interface.

Documents and Deliverables

These are the related documents and deliverables of this project.

About Us

This is a individual project under the supervision of one supervisor:

Image

Chi Ian Tang

Project Leader

Chi Ian Tang (Michael) is currently in his final year at the Department of Computer Science, the University of Hong Kong.

Image

Dr. S. M. Yiu

Project Supervisor

Get In Touch

For enquiries about the current status of the project, please get in contact via the following channels.

Contact Info

  • citang@cs.hku.hk
  • u3520924@connect.hku.hk