Effect of User-Prompted Priors on Semi-Automated Cancer Lesion Segmentation in Whole-Body Computed Tomography
2026-07-27 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors studied how to better measure cancer spread in body scans by using computer programs that help outline tumors. Normally, doctors either measure only a few spots or manually mark all tumors, which is slow and hard. They found that if users give the program simple hints, like drawing shapes on three different views of the body, the computer can more accurately identify tumors automatically. This method worked much better than when no hints were given, making it easier to get good 3D tumor data. Their work helps speed up and improve cancer volume measurements.
Metastatic cancerRECISTTotal tumour volumeSegmentationComputed tomographyDice scoreUser-prompted priorsSemi-automated segmentationOrthogonal planesGround-truth data
Authors
Isac Stark, Johan Öfverstedt, Elin Lundström, Simon Ekström, Håkan Ahlström, Joel Kullberg
Abstract
In clinical oncology studies, metastatic cancer is commonly evaluated using "Response Evaluation Criteria in Solid Tumors" (RECIST), in which the diameter of up to five lesions is measured and followed over the course of treatment. However, RECIST shows limited correlation with overall survival. Total tumour volume (TTV) is a stronger predictor but typically relies on manual ground-truth segmentation of all lesions, which is time-consuming and requires expert domain knowledge. Semi-automated approaches leveraging user-prompted priors, such as bounding boxes and single-slice contours, as inputs to automated segmentation methods can facilitate the generation of ground-truth segmentations. This work investigates the impact of different user-prompted priors on semi-automated cancer lesion segmentation performance in whole-body computed tomography. Across 3-fold cross-validation and external testing, more complex spatial priors consistently improved performance, with contour priors from three orthogonal planes (axial, coronal and sagittal) achieving the best results. On the external test (n=3865 lesions), this approach achieved a mean Dice score of 0.882, compared to a mean Dice score of 0.671 for the baseline model with no spatial prior. These findings suggest that the use of multi-plane orthogonal user-prompted priors can improve semi-automated tumour lesion segmentation and support efficient generation of high-quality volumetric ground-truth data.