Duygu Atasoy | Searcharxiv

arXiv2026

Anchored Fusion for Automated Airway Segmentation in Pediatric and Adult CT

Airway segmentation from chest CT enables branch-level measurement of airway size and growth, but peripheral bronchi remain hard to segment, particularly in children, whose airways are smaller and for whom few annotated scans exist. Different models recover complementary branches of the airway tree. Their union improves overall tree coverage but also introduces disconnected non-airway objects. We propose anchored fusion (A-Fuse), which takes the voxel-wise union of TfeNet, a retrained MEDPSeg and a reimplemented AirMorph and retains the connected component with greatest overlap with the AirMorph mask. On 30 held-out adult CTs from the ATM'26 training set, A-Fuse reaches a Dice score of 93.24% and the highest branch detection rate among compared methods (92.05%). On nine pediatric scans, it reaches 94.15% Dice and 68.79% branch detection. A-Fuse retains the airway tree in all 39 test scans, whereas keeping the largest component selects a non-airway object in four. A-Fuse is deployed as an automated web application and we release all code, trained weights and an open-source ITK-SNAP extension for 3D tubular segmentation editing.

eess.IV↗

arXiv2026

WristMIR: Coarse-to-Fine Region-Aware Retrieval of Pediatric Wrist Radiographs with Radiology Report-Driven Learning

Retrieving wrist radiographs with analogous fracture patterns is challenging because clinically important cues are subtle, highly localized and often obscured by overlapping anatomy or variable imaging views. Progress is further limited by the scarcity of large, well-annotated datasets for case-based medical image retrieval. We introduce WristMIR, a region-aware pediatric wrist radiograph retrieval framework that leverages dense radiology reports and bone-specific localization to learn fine-grained, clinically meaningful image representations without any manual image-level annotations. Using MedGemma-based structured report mining to generate both global and region-level captions, together with pre-processed wrist images and bone-specific crops of the distal radius, distal ulna, and ulnar styloid, WristMIR jointly trains global and local contrastive encoders and performs a two-stage retrieval process: (1) coarse global matching to identify candidate exams, followed by (2) region-conditioned reranking aligned to a predefined anatomical bone region. WristMIR improves retrieval performance over strong vision-language baselines, raising image-to-text Recall@5 from 0.82% to 9.35%. Its embeddings also yield stronger fracture classification (AUROC 0.949, AUPRC 0.953). In region-aware evaluation, the two-stage design markedly improves retrieval-based fracture diagnosis, increasing mean $F_1$ from 0.568 to 0.753, and radiologists rate its retrieved cases as more clinically relevant, with mean scores rising from 3.36 to 4.35. These findings highlight the potential of anatomically guided retrieval to enhance diagnostic reasoning and support clinical decision-making in pediatric musculoskeletal imaging. The source code is publicly available at https://github.com/quin-med-harvard-edu/WristMIR.

cs.CV↗