Celebrating the defence
Stine Hansen defended her PhD thesis “Leveraging Supervoxels for Medical Image Volume Segmentation With Limited Supervision” on Dec. 16th at UiT The Arctic University of Norway.
The focus of Stine Hansens thesis is on the development of new ML algorithms for medical image segmentation with limited supervision. In particular, completely unsupervised solutions to lung tumor segmentation and solutions that only require a few labeled samples to perform organ segmentation (few-shot learning) are proposed. To exploit the data-specific opportunities of medical images, automatically generated supervoxels are exploited in various ways to solve key challenges related to the task.
.jpg)
Evaluation Committee
Supervisors
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.
Last week, we wished Aasland—accompanied by Political Advisor Munir Jaber and Senior Adviser Finn-Hugo Markussen—welcome to the Norwegian Computing Center (NR). One of the visit's goals was to showcase ongoing Visual Intelligence projects at NR.
Alba Ordoñez and Anders U. Waldeland presented ongoing work on seismic foundation models and an interactive seismic interpretation engine at EAGE Annual 2025 in Toulouse, France.
Østmo was invited to Abels tårn—one of the largest popular science radio shows in Norway—to answer listener-submitted questions related to artificial Intelligence (AI). The live show took place at Blårock Cafe in Tromsø, Norway on June 12th.
VI industry partner Kongsberg Satellite Services (KSAT) received an Industrial PhD grant from the Research Council of Norway. The project will be closely connected to Visual Intelligence's "Earth observation" innovation area.
Visual Intelligence Researchers Amund Vedal and Arnt Børre Salberg recently presented ongoing Visual Intelligence research at a plankton-themed workshop organized by the Institute of Marine Research (IMR), Norway
We happily welcome Dennis Adamek, who recently started as a Doctoral Research Fellow at Visual Intelligence in Tromsø in April 2025.
Congratulations to Professor Lars Ailo for receiving the NORA Award for Diversity in AI at the NORA Annual Conference in Halden, Norway.
Teresa Dorszewski is a PhD Candidate at the Section for Cognitive Systems at the Technical University of Denmark. She visited Visual Intelligence in Tromsø from January to April 2025.
Centre Director Robert Jenssen was invited to give a keynote and participate in a panel discussion on AI as critical national infrastructure at the NORA Annual Conference 2025 in Halden, Norway.
Både forskere og statsråd mener kunstig intelligens bør tydelig merkes. Men forlaget som lager puslespillet som ekspertene mener er KI-generert, sier de ikke har noe med hvordan illustratører lager produktene sine (Norwegian news article by NRK)
Researchers have developed an artificial intelligence to map specific immune cells in lung cancer tumors. It can lead to less costly examinations and more personalised cancer treatment (English news story on sciencenorway.no).
The Department of Physics and Technology at UiT The Arctic University of Norway is pleased to announce 4 exciting PhD Fellowships within machine learning at SFI Visual Intelligence. Application deadline: June 17th.
.jpg)
Norsk forskning har utviklet kunstig intelligens som raskt kan analysere lungekreft. Ekspertene forklarer hvordan dette kan bidra til en mer effektiv og persontilpasset behandling (Norwegian news article in vg.no)
We invite submissions presenting new and original research on all aspects of Deep Learning for the Northern Lights Deep Learning (NLDL) Conference 2026.
Partiet Sentrum er blant dem som tar grepet. – Ganske smart, mener valgforsker (Norwegian news article in TV2.no)
Visual Intelligence will be well presented at ICML 2025—one of the leading international academic conference in machine learning—with four newly accepted research papers.
After four years of operation, an international AI expert panel was appointed to assess Visual Intelligence's progress and results. The evaluation was characterized by its excellent remarks on the centre's scientific quality and innovation output.
Senior Researcher Fredrik Dahl recently gave a talk about Norsk Regnesentral's work on developing AI algorithms for automatic analysis of image quality and cancer detection at Norsk Radiografforbund's mammography symposium in Oslo.
Han høster anerkjennelse for sitt forskningsarbeid innen utvikling av mer effektive KI-modeller (Norwegian news article in Nordlys).
.jpg)
Congratulations to associate professor Ali Ramezani-Kebrya for receiving a prestigious FRIPRO grant from the Research Council of Norway!