Cross-Source Supervision for Bone Infection Segmentation in Dual-Modality PET-CT

2026-05-10

AI Summary

This study addresses the challenge of segmenting bone infections in PET-CT images, where ambiguous lesion boundaries and inter-expert annotation inconsistencies hinder reliable delineation. To overcome this, the authors propose an end-to-end dual-modality segmentation framework that fuses PET metabolic and CT bone-window anatomical information at an early stage. A decoupled dual-source supervision strategy is introduced, enabling parallel training on independent expert annotations aligned with distinct clinical intents—high sensitivity versus high specificity—thereby avoiding forced consensus. Patient-level 3D voxel evaluation combined with rigorous cross-validation mitigates slice-wise correlation bias under limited sample sizes. Experimental results demonstrate that the model effectively internalizes divergent diagnostic philosophies, significantly enhancing segmentation robustness and clinical applicability at the patient level (reported as mean ± standard deviation).

Abstract

Early and accurate diagnosis and lesion localization of bone infections are crucial for clinical treatment. PET-CT integrates anatomical information from CT with metabolic information from PET, making it an important imaging modality for diagnosing bone infections. However, accurate lesion segmentation remains challenging due to indistinct lesion boundaries and inconsistencies in annotations generated by different experts or automated systems. In this work, we investigate multimodal segmentation of bone infections under annotation discrepancy. We develop a bimodal end-to-end segmentation framework that integrates PET metabolic signals and CT bone-window anatomy through an early-fusion multimodal representation.To mitigate performance inflation caused by inter-slice correlation in small datasets, this study discards traditional two-dimensional evaluation methods and implements a rigorous patient-level 3D volumetric evaluation and cross-validation. Furthermore, instead of forcing a singular consensus, we propose a decoupled dual-source learning framework where parallel models are trained on independent expert annotations driven by high-sensitivity and high-specificity clinical intents. Experimental results objectively report performance variations at the patient level (Mean + SD and Mean - SD), demonstrating the effectiveness of multimodal PET-CT fusion. The cross-evaluation matrix quantitatively reveals how models successfully internalize distinct expert diagnostic philosophies, providing a robust, diversity-preserving paradigm for clinical AI deployment in bone infection segmentation.

Problem

• bone infection segmentation
• PET-CT
• annotation discrepancy
• multimodal imaging
• lesion localization

Innovation

• multimodal fusion
• cross-source supervision
• 3D volumetric evaluation
• decoupled dual-source learning
• bone infection segmentation