Submit before June 30 to help us manage compute load & be eligible for the Early Bird prizes!
T14: Classifying kidney abnormality¶
Objective:
This task aims to identify the presence of significant kidney abnormalities based on the radiology report.
Patient Population:
Radiology reports (in Dutch) were collected from patients who underwent CT scans of the kidney at Radboudumc.
Imaging Data:
Not applicable. The task is based solely on textual data — radiology reports written in Dutch.
Test Data:
The test set consists of 183 cases.
Reference Standard:
Reports were manually analyzed and labeled as either "normal" or "abnormal" by a native Dutch-speaking PhD candidate.
Evaluation Metrics:
The classification performance will be evaluated using the AUC.
Relation to Existing Challenges:¶
- Task 14 is derived from Task003 (Kidney abnormality identification) of the DRAGON challenge.
- Unlike the DRAGON challenge, UNICORN allows training on platform using only a small set of few-shot examples per task.