Development of diagnosis support system for endoscopic images using machine learning

Gastrointestinal (GI) cancers are the leading cause of cancer worldwide. Early detection of abnormalities in the GI tract can reduce the chances of GI cancers and provide successful treatment. Towards fulfilling this goal, we target to build novel generalizable and robust machine learning algorithms that can potentially find the miss-detected lesion that were left out during manual endoscopy and improve the healthcare system by automatically detecting, segmenting, and localizing diseases and other endoscopic findings inside the GI tract.
Master

Colonoscopy is the gold standard for colon examinations. The factors such as gaps among endoscopists skill and experience and quality of bowel preparation can lead to the miss-detection of the lesions. The other challenging factors are related to the structure and characteristics of lesions (for example, size, color, shape, and occurrence), making the detection more difficult. One of the solutions could be the use of machine learning to detect such diseases. However, machine learning algorithms require many samples for the training and development of the algorithms. In this respect, we will develop a new dataset with the help of medical experts and make it publicly available to the multimedia community. Later on, we will use the same dataset and develop new algorithms and provide a new benchmark.

Goal

To develop a fully automated system for endoscopic disease segmentation, detection, and localization

Learning outcome

Deep understanding of endoscopic image analysis
Working on a real-world application
Possibility of collaboration with researchers
Opportunity to implement and research a novel approach
Opportunity to participate in challenges and conferences

Qualifications

Experience with Python programming
Understanding of deep learning will be a benefit.

Supervisors

  • Pål Halvorsen
  • Michael Riegler
  • Debesh Jha

 

Collaboration partners

Simula Metropolitan Center For Digital Engineering AS

References

  1. D. Jha, M. A. Riegler, D. Johansen, P. Halvorsen, and H. D. Johansen“ DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation,” CBMS, 2020.
  2. Sharib Ali et al., “An objective comparison of detection and segmentation algorithms for artefacts in clinical endoscopy”, Scientific Reports, 2020.

Contact person