CACTUS: Multiview classifier for Punctate White Matter Lesions detection & segmentation in cranial ultrasound volumes

• Published in: Computers in biology and medicine Vol. 197; no. Pt B; p. 111085
• Main Authors: Estermann, Flora, Kaftandjian, Valerie, Guy, Philippe, Quetin, Philippe, Delachartre, Philippe
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.10.2025; Elsevier
• Subjects: 3D Ultrasound; Acoustics; Algorithms; Brain Lesions; Computer Science; Cross-Attention; Deep Learning; Engineering Sciences; Female; Humans; Image Interpretation, Computer-Assisted - methods; Image Processing, Computer-Assisted - methods; Infant, Newborn; Internal Medicine; Life Sciences; Machine Learning; Magnetic Resonance Imaging; Male; Medical Imaging; Neurons and Cognition; Other; Ultrasonography - methods; Vision Transformers; White Matter - diagnostic imaging; Brain Lesions; Vision Transformers; Cross-Attention; 3D Ultrasound; Deep Learning
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2025.111085

Punctate white matter lesions (PWML) are the most common white matter injuries found in preterm neonates, with several studies indicating a connection between these lesions and negative long-term outcomes. Automated detection of PWML through ultrasound (US) imaging could assist doctors in diagnosis more effectively and at a lower cost than MRI. However, this task is highly challenging because of the lesions’ small size and low contrast, and the number of lesions can vary significantly between subjects. In this work, we propose a two-phase approach: (1) Segmentation using a vision transformer to increase the detection rate of lesions. (2) Multi-view classification leveraging cross-attention to reduce false positives and enhance precision. We also investigate multiple postprocessing approaches to ensure prediction quality and compare our results with what is known in MRI. Our method demonstrates improved performance in PWML detection on US images, achieving recall and precision rates of 0.84 and 0.70, respectively, representing an increase of 2% and 10% over the best published US models. Moreover, by reducing the task to a slightly simpler problem (detection of MRI-visible PWML), the model achieves 0.82 recall and 0.89 precision, which is equivalent to the latest method in MRI. [Display omitted] •PWML are very small brain lesions occurring in 18%–35% of preterm infants.•Automatic detection and segmentation of PWML is possible through ultrasound imaging.•The use of multiple brain projections gives the model more spatial context.•Multi-view classification using cross-attention significantly improves accuracy.•Segmentation followed by multi-view classification works well for 3D volumes.