Undergraduate studies in Machine learning for early prediction of primary sclerosing cholangitis (scholarship)

Do you want to contribute to improving human health?

Primary sclerosing cholangitis (PSC) is a chronic liver disease that often occurs together with inflammatory bowel disease (IBD) and is typically diagnosed late. Our lab use linked Swedish national health registers to develop machine-learning models that aim to identify, five years before PSC is diagnosed, which patients with IBD will develop the disease. As a scholarship student, you will work on this model with nationwide register data and learn how clinical prediction models are validated, explained and reported. Our team is multidisciplinary with clinical and methodological expertise in the field.

Division

The project is based in the Gastroenterology Unit at the Department of Medicine, Huddinge, Karolinska Institutet, Campus Flemingsberg. It is led by Ghada Nouairia, Assistant Professor in Computational Oncology, whose research focuses on PSC, cholangiocarcinoma and gallbladder cancer, and on clinically applicable prediction models that integrate registry and electronic health record data, medical imaging and molecular biomarkers.

The study links the National Patient Register, the Prescribed Drug Register, the Cause of Death Register, the Swedish Cancer Register and population data from Statistics Sweden to follow patients with IBD from their first healthcare contacts until a PSC diagnosis. The work is carried out together with hepatologists and a biostatistician.

Duties (of the scholarship student)

You will take part in the validation and reporting of the PSC prediction model. The tasks will be set out in an individual research plan, adapted to the length of the scholarship period, and may include:

• Data quality and cohort description: checking that every predictor is derived only from information recorded before the prediction time point, documenting how each feature was constructed, and preparing descriptive tables of the study cohort.
• Validation analyses in R or Python, such as bootstrap confidence intervals, calibration of predicted risks, and model performance across age groups and at different times before diagnosis.
• Figures, including receiver operating characteristic (ROC) and precision–recall curves, enrichment curves showing the share of future PSC cases captured, and model explanations based on SHapley Additive exPlanations (SHAP).
• A structured literature review on PSC in IBD and on existing PSC prediction models, and mapping the study against the TRIPOD+AI reporting guideline for clinical prediction models.
• Writing documented, reproducible analysis code.

Eligibility requirements

These scholarships aim to provide students at the undergraduate level with early practical experience of research work, in order to strengthen the link between education and research and encourage future studies at the doctoral level.

The scholarships can be granted either for studies in an existing course or for a shorter period of practical research experience (documented in an individual study plan/research plan) for admitted students at the undergraduate or graduate level. The scholarships may be awarded for up to 12 months (divided into a maximum of four courses/periods) during the study period. Decisions on scholarships may be made for a maximum of six months at a time.

The scholarships may be granted only to: persons who have been admitted (and registered in Ladok) to KI to study at the undergraduate or graduate level, persons who are admitted (and registered) to undergraduate or graduate level programs at a university with which KI collaborates.

Skills and personal qualities

Required:

• Studies in data science, machine learning engineering or a closely related field.
• Good programming skills in R and Python.
• Clear communication and documentation skills to be able to explain and replicate research findings.
• Good timekeeping skills and an ability to work in a team or independently.

Meritious:

• Earlier experience with registry data.
• Experience or knowledge on PSC and/or biliary tract cancers.

What do we offer?

A creative and inspiring environment with wide-ranging expertise and interests. Karolinska Institutet is one of the world's leading medical universities. Here, we conduct innovative medical research and provide the largest range of biomedical education in Sweden. At KI, you get to meet researchers working with a wide range of specialisms and methods, giving you ample opportunity to exchange knowledge and experience with the various scientific fields within medicine and health. It is the crossover collaborations, which have pushed KI to where it is today, at the forefront of global research. Several of the people you meet in healthcare are educated at KI. A close relationship with the health care providers is important for creating groundbreaking top quality education and research. Students and employees have free access to our modern gym facilities with trained staff on site.

Location: Flemingsberg

Application process

AnapplicationmustcontainthefollowingdocumentsinEnglish*:*

• A complete curriculum vitae,
• A summary of current work (no more than one page)

The application is to be submitted on the Varbi recruitment system.

Want to make a difference? Join us and contribute to better health for all

Anställningsform: scholarship | Anställningens omfattning: heltid | Antal lediga befattningar: 1 | Sysselsättningsgrad: 100 | Ort: Huddinge | Län: Stockholms län | Land: Sweden | Referensnummer: STÖD 2-3751/2026 | Kontakt: Ghada Nouairia , Karin Gåse , | Facklig företrädare: Mari Gilljam, OFR , Peter Svensson, SACO , Robert Lindberg, SEKO , | Publicerat: 2026-09-16 | Sista ansökningsdag: 2026-10-07