Stine Hansen

Title of the PhD thesis:

"Leveraging Supervoxels for Medical Image Volume Segmentation With Limited Supervision"

Abstract

The majority of existing methods for machine learning-based medical image segmentation are supervised models that require large amounts of fully annotated images. These types of datasets are typically not available in the medical domain and are difficult and expensive to generate. A wide-spread use of machine learning based models for medical image segmentation therefore requires the development of data-efficient algorithms that only require limited supervision. To address these challenges, this thesis presents new machine learning methodology for unsupervised lung tumor segmentation and few-shot learning based organ segmentation. When working in the limited supervision paradigm, exploiting the available information in the data is key. The methodology developed in this thesis leverages automatically generated supervoxels in various ways to exploit the structural information in the images. The work on unsupervised tumor segmentation explores the opportunity of performing clustering on a population-level in order to provide the algorithm with as much information as possible. To facilitate this population-level across-patient clustering, supervoxel representations are exploited to reduce the number of samples, and thereby the computational cost. In the work on few-shot learning-based organ segmentation, supervoxels are used to generate pseudo-labels for self-supervised training. Further, to obtain a model that is robust to the typically large and inhomogeneous background class, a novel anomaly detection-inspired classifier is proposed to ease the modelling of the background. To encourage the resulting segmentation maps to respect edges defined in the input space, a supervoxel-informed feature refinement module is proposed to refine the embedded feature vectors during inference. Finally, to improve trustworthiness, an architecture-agnostic mechanism to estimate model uncertainty in few-shot segmentation is developed. Results demonstrate that supervoxels are versatile tools for leveraging structural information in medical data when training segmentation models with limited supervision.

The thesis is published and available in Munin

Supervisors

• Professor Robert Jenssen, IFT, UiT (main supervisor)
• Associate professor Stian Normann Anfinsen, IFT, UiT (co-supervisor)

Evaluation Committee

• Professor Alexandros Iosifidis, Department of Electrical and Computer Engineering, Aarhus University,
  Danmark (1. Opponent)
• Professor Kjersti Engan, Institutt for data og teknologi, Universitetet i Stavanger (2. Opponent)
• Associate professor Shujian Yu, IFT, UiT (internal member)

Links to the trial lecture and defense will be possible to open when the live stream begins. If you haven`t clicked the link to the folder before it begins, refresh the web browser for them to become visible. If you have clicked the link to the trial lecture or defense before it has started, it will open automatically when the stream begins. 

The trial lecture starts at 10.15 December 16th

Trial lecture

The defense starts at 12.15 December 16th

Defense