From structural complexity to causal representation: A dynamic fractal-attention framework for fine-grained ovarian tumor classification in ultrasound

Qingyuan Zhang; Qiuyue Fu; Xuping Zhang; Xuecheng Wang; Chidao Chen; Zhao Zhang; Peng Wang

doi:10.1016/j.media.2026.104101

From structural complexity to causal representation: A dynamic fractal-attention framework for fine-grained ovarian tumor classification in ultrasound

Med Image Anal. 2026 Apr 24:112:104101. doi: 10.1016/j.media.2026.104101. Online ahead of print.

Authors

Qingyuan Zhang¹, Qiuyue Fu², Xuping Zhang³, Xuecheng Wang³, Chidao Chen⁴, Zhao Zhang⁴, Peng Wang⁵

Affiliations

¹ Department of Ultrasonography, Tengzhou Central People's Hospital, Tengzhou, 277500, China.
² Sichuan Provincial Women's and Children's Hospital/The Affiliated Women's and Children's Hospital of Chengdu Medical College, Chengdu, 610000, China.
³ Department of Medical Ultrasound, Shandong Medicine and Health Key Laboratory of Abdominal Medical Imaging, The First Affiliated Hospital of Shandong First Medical University & Shandong Provincial Qianfoshan Hospital, Jinan, 250014, China.
⁴ School of Medical Imaging, Qilu Medical University, Zibo, 255300, China.
⁵ School of Medical Imaging, Qilu Medical University, Zibo, 255300, China. Electronic address: wangpeng9138@163.com.

PMID: 42035560
DOI: 10.1016/j.media.2026.104101

Abstract

Accurate fine-grained classification of ovarian tumors from ultrasound images remains challenging due to speckle noise, boundary ambiguity, structural heterogeneity, and acquisition-induced spurious correlations. To address these issues, we propose a structure-aware and causally regularized deep learning framework that integrates dynamically learnable fractal modeling, edge-guided structural refinement, and a causal inference-assisted module within a unified architecture. The framework explicitly models lesion-intrinsic structural complexity while suppressing non-discriminative acquisition cues, enabling robust and interpretable subtype recognition. Extensive experiments on a large private ovarian ultrasound cohort and the public MMOTU benchmark demonstrate that the proposed method outperforms representative CNNs, Transformer-based models, self-supervised approaches, and established causal baselines, achieving higher accuracy and macro-F1 with consistently narrower confidence intervals. Patient-level evaluation further shows that dynamic fractal modeling captures disease-intrinsic structural complexity with cross-view stability, leading to substantial improvements in recall and F1-score for structurally complex and low-prevalence subtypes. Comprehensive interpretability analyses, including quantitative structural saliency assessment and Grad-CAM visualization, confirm that predictions are primarily driven by clinically meaningful intralesional structures rather than boundary contrast or acquisition artifacts. A retrospective reader study further demonstrates that AI assistance improves diagnostic consistency and reduces systematic misclassification, particularly among less experienced sonographers. Overall, this work establishes a theoretically grounded and empirically validated framework for consistent and interpretable fine-grained ovarian tumor classification from ultrasound imaging.

Keywords: Causal inference; Fractal feature; Ovarian tumor; Ultrasound image.