Projects at ICGI

Learning from Deep Learning

Deep learning allows computers to do what comes naturally to humans: learn by example. Increasing knowledge of the biological basis for the predictions made by computerized systems is essential for implementation in the clinic.


From bench to bedside

Developing systems with straight forward clinical utility has been a critical goal for the research activity in DoMore!. The focal point has been to bring methods and products to a level where they can be easily implemented in the clinic and to establish the structures and collaborations that enable efficient commercialisation


A variety of factors influence a patient’s clinical outcome, including intrinsic characteristics of the patient, disease, or medical condition, and the effects of any treatments that the patient receives. This project’s primary purpose is to develop dedicated predictive markers for specific clinical treatment regimens using convolutional neural networks (deep learning) combined with our prognostic biomarkers.

Polyp classification

Due to the shortage of endoscopists and pathologists with the required experience, automated risk classification of colon polyps will significantly reduce the workload. Automated systems will also improve objectivity, and allow for the increased throughput needed to enable a sucessful bowel cancer screening program. 

The Nucleolus

The nucleolus is the most prominent structure in a cell nucleus. It is the site of ribosomal RNA (rRNA) transcription, pre-rRNA processing and ribosome subunit assembly. The nucleolus is a dynamic structure that assembles around the clusters of rRNA gene repeats during late telophase, persists throughout interphase and then disassembles as cells enter mitosis. Alterations in both number and shape of nucleoli are linked to cancer and cancer prognosis.

mRNA DNA in situ hybridization

In situ hybridization (ISH) uses a labelled probe (complementary DNA, RNA or modified nucleic acids strand) to localize a specific DNA or RNA sequence in a portion or section of the tissue - in situ. DNA ISH can be used to determine the structure of chromosomes, and RNA ISH can be used to measure and localize mRNAs within individual cells in tissue sections.

mRNA Expression of Candidate Genes

Cancer is a genetic disease, that is, cancer is caused by certain changes to genes that control cell functions, especially how they grow and divide. These changes must be expressed (or influence expression of other genes) to cause damage to cells. .

SeeMore - a solution for viewing virtual slides and integrated analysis

In most of our projects, we scan several slides for every case in the project. In order to view these scans in the best way possible, we need an easy to use, fast and general application that works for scans from different scanner manufacturers. There are several commercially available viewers, but they are missing important features for visualising results. To meet the demands of current and future projects, we are now building a new scanner called SeeMore!

HighRes - The High resolution scanner project

Commercially available scanners are designed for speed and image viewing, not for image analysis. Hence, the image resolution is inferior to standard light microscopy and with extremely limited flexibility to control or alter the light. The High-Res project aims to create a flexible and autonomous system that will push the boundaries of histopathology imaging. 

Mitotic index

Mitotic Index (MI) is defined as the ratio between the number of cells in a population undergoing mitosis to the total number of cells in a population.

In silico Pathology

The digitalisation of pathology is currently being established in most pathology departments nationally and internationally, and the digitalisation of routine sections poses many advantages in both diagnosis and prognosis.


Developing methods using artificial intelligence (AI) to give cancer patients a more precise prognosis and counteract overtreatment.

Focus classification

Successful imaging depends on focus. Although our scanners and microscopes are equipped with autofocus algorithms, they frequently produce images that are partially or totally out of focus. Manual control detects many of these cases, but the manual process of checking is not only resource-demanding but also incomplete. 


Image segmentation is the process of partitioning a digital image into multiple segments, sets of pixels, also known as image objects. The purpose of segmentation is to simplify or change an image’s representation into something more meaningful and easier to analyse.

Active surveillance (prostate cancer)

Prostate cancer ranks as the most common cancer among Norwegian men, with about 5000 new diagnoses yearly. The high incidence of overdiagnosis and overtreatment—impacting nearly two-thirds of cases—underscores the critical need for a more accurate risk stratification system. Our project aims to address this challenge by developing a precise risk assessment tool utilizing artificial intelligence (AI) biomarkers applicable to biopsy specimens.

Tumour heterogeneity

Cancer is a dynamic disease. Over the course of the disease, cancers generally become more heterogeneous. Tumour heterogeneity describes the observation that different tumour cells can show distinct morphological and phenotypic profiles, including cellular morphology, gene expression, metabolism, motility, proliferation, and metastatic potential, all with potentially differential levels of sensitivity to treatment.

Protein Expression

Dysregulation of different genes is known to contribute to cancer development through several different mechanisms. Divergent levels of a protein may provide information relevant to diagnosis, prognosis or treatment decisions for patients.

Cytogenetic and Molecular Analysis of Female Genital Tract Tumours

Cancer of the female genital tract is the third most common group of malignancies in women, exceeded in frequency only by cancer of the breast and digestive tract. Most of the cancers are of the ovaries and uterus, but tumors also occur in the fallopian tubes, vulva, and vagina. In recent decades, genetic analysis of tumor cells has shed considerable light on the mechanisms of tumorigenesis and is increasingly relied upon to provide prognostic and diagnostic information about cancer diseases.

Kreftlex is a digital encyclopedia on cancer and cancer treatment developed and maintained by the Institute for Cancer Genetics and Informatics at Oslo University Hospital.

Oncolex: a web-based Cancer Encyclopedia

Oncolex is a web based cancer encyclopedia for health care providers worldwide, published by the Institut for Cancer genetics and Informatics (former Institute for Medical Informatics (IMI)) at Oslo University Hospital. IMI is part of the Center for Cancer Biomedicine.


The relation between chromatin organisation characteristics and cancer prognosis has been one of our main research projects since the establishment of the Institute in 2004. Chromatin is a complex of DNA and histones that are packaged into thin fibres within the nucleus of eukaryotic cells.


– a conceptual framework to provide standardized development of medical registries, developed by Håvard E. G. Danielsen and the Institute for Medical Informatics (Institute for Cancer Genetics and Informatics). 

DNA Ploidy

Chromosome instability, either past or present, is indicated whenever tumour cells harbour an abnormal quantity of DNA, termed ‘aneuploidy’. Abnormalities of cellular DNA content (polyploidy and aneuploidy) have long been associated with tumorigenesis. 

Aneuploidy - Promoter and suppressor of malignant grouwth

Aneuploidy is a very rare and tissue-specific event in normal conditions, occurring in a low number of brain and liver cells. Its frequency increases in age-related disorders and has been recognised as a hallmark of tumorigenesis for more than a century. Still, the mechanism behind chromosomal errors and malignant growth has remained obscure.