A public cardiac CT dataset featuring the left atrial appendage

2025-10-07

AI Summary

Accurate segmentation of the left atrial appendage (LAA), coronary arteries (CAs), and pulmonary veins (PVs) in cardiac CT remains challenging due to severe annotation scarcity and anatomical complexity. To address this, we introduce the first open-source, high-resolution whole-heart CT dataset comprising 1,000 cases, with expert-level, fine-grained segmentation masks for the LAA, CAs, PVs, and entire heart, alongside a systematically curated catalog of common scanning artifacts. We propose an anatomy-consistency-driven annotation refinement pipeline that integrates expert manual correction, TotalSegmentator-based pre-segmentation, and refinement of ImageCAS raw annotations. Models are trained using a state-of-the-art segmentation framework specifically optimized for high-resolution CT. This dataset substantially improves reliability in LAA morphological analysis and achieves superior segmentation accuracy for CAs and PVs. It serves as a critical open resource for algorithm development, robustness evaluation, and clinical translation.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Large Multimodal Models (LMMs)
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Representation, semantic annotation, enhancement, enrichments, access and/or integration of a variety of data on the Web
• Security and Privacy: Data transparency and provenance

Abstract

Despite the success of advanced segmentation frameworks such as TotalSegmentator (TS), accurate segmentations of the left atrial appendage (LAA), coronary arteries (CAs), and pulmonary veins (PVs) remain a significant challenge in medical imaging. In this work, we present the first open-source, anatomically coherent dataset of curated, high-resolution segmentations for these structures, supplemented with whole-heart labels produced by TS on the publicly available ImageCAS dataset consisting of 1000 cardiac computed tomography angiography (CCTA) scans. One purpose of the data set is to foster novel approaches to the analysis of LAA morphology. LAA segmentations on ImageCAS were generated using a state-of-the-art segmentation framework developed specifically for high resolution LAA segmentation. We trained the network on a large private dataset with manual annotations provided by medical readers guided by a trained cardiologist and transferred the model to ImageCAS data. CA labels were improved from the original ImageCAS annotations, while PV segmentations were refined from TS outputs. In addition, we provide a list of scans from ImageCAS that contains common data flaws such as step artefacts, LAAs extending beyond the scanner's field of view, and other types of data defects.

Problem

• Accurate segmentation of cardiac structures remains challenging in medical imaging
• Creating open-source dataset for left atrial appendage and coronary arteries
• Providing curated cardiac CT scans with common data flaws for analysis

Innovation

• Created open-source cardiac CT dataset with curated segmentations
• Used specialized segmentation framework for LAA structure analysis
• Provided annotated data with common imaging artifacts identification