Toward Content-based Indexing and Retrieval of Head and Neck CT with Abscess Segmentation

2025-12-01

AI Summary

This study addresses the challenges of automatic abscess segmentation and content-aware retrieval in head-and-neck CT images. We introduce AbscessHeNe, the first large-scale, comprehensively annotated dataset (4,926 slices) featuring pixel-level segmentation masks and rich clinical metadata—enabling both semantic segmentation modeling and deep neck space involvement assessment. We propose a multi-architecture segmentation framework integrating CNN, Transformer, and Mamba backbones, evaluated using Dice, IoU, and normalized surface distance (NSD). The best-performing model achieves Dice=0.39, IoU=0.27, and NSD=0.67, underscoring the task’s high complexity. AbscessHeNe is the first dataset explicitly designed for abscess-centric content indexing and case-based retrieval in medical imaging, filling a critical gap in the field. It will be publicly released to advance intelligent diagnosis and cross-case knowledge discovery.

Technology Category

• Computer Vision: Segmentation
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Abscesses in the head and neck represent an acute infectious process that can potentially lead to sepsis or mortality if not diagnosed and managed promptly. Accurate detection and delineation of these lesions on imaging are essential for diagnosis, treatment planning, and surgical intervention. In this study, we introduce AbscessHeNe, a curated and comprehensively annotated dataset comprising 4,926 contrast-enhanced CT slices with clinically confirmed head and neck abscesses. The dataset is designed to facilitate the development of robust semantic segmentation models that can accurately delineate abscess boundaries and evaluate deep neck space involvement, thereby supporting informed clinical decision-making. To establish performance baselines, we evaluate several state-of-the-art segmentation architectures, including CNN, Transformer, and Mamba-based models. The highest-performing model achieved a Dice Similarity Coefficient of 0.39, Intersection-over-Union of 0.27, and Normalized Surface Distance of 0.67, indicating the challenges of this task and the need for further research. Beyond segmentation, AbscessHeNe is structured for future applications in content-based multimedia indexing and case-based retrieval. Each CT scan is linked with pixel-level annotations and clinical metadata, providing a foundation for building intelligent retrieval systems and supporting knowledge-driven clinical workflows. The dataset will be made publicly available at https://github.com/drthaodao3101/AbscessHeNe.git.

Problem

• Develop a dataset for head and neck abscess segmentation in CT scans
• Evaluate segmentation models to delineate abscess boundaries accurately
• Enable content-based indexing and retrieval for clinical decision support

Innovation

• Created a curated dataset with pixel-level abscess annotations
• Evaluated CNN, Transformer, and Mamba models for segmentation
• Designed dataset for content-based indexing and retrieval systems