Development and Validation of AI System for AAS CTA Diagnosis and Mapping
收藏资源简介:
AAS-DSS: AI-Powered Acute Aortic Syndrome Decision Support System Code This repository contains the code for AAS-DSS, a Python-based pipeline for automated detection/classification of acute aortic syndrome (AAS) subtypes (AD, IMH, PAU) using CTA images. Core Functions: - Preprocessing & aorta mask generation via TotalSegmentator. - Multi-class segmentation (AD/IMH/PAU) with nnUNet (requires pre-trained weights). - Slice-level classification using custom .pth models. - User-friendly GUI for path configuration and pipeline execution. Prerequisites: Python 3.x, medical imaging libraries, TotalSegmentator, pre-trained model weights (per specified directories). Input/Output: Accepts .nii.gz CTA images; outputs results to a user-specified folder. Flexible path adjustment supported. AAS-DSS enables automated 17-zone aortic segmentation (extended SVS/STS classification) to support standardized AAS diagnosis.
AAS-DSS:人工智能驱动的急性主动脉综合征决策支持系统代码 本仓库包含AAS-DSS的相关代码,该系统是一款基于Python的工作流,可利用计算机断层血管造影(Computed Tomographic Angiography,CTA)图像自动检测并分类急性主动脉综合征(Acute Aortic Syndrome,AAS)亚型,包括主动脉夹层(Aortic Dissection,AD)、主动脉壁内血肿(Intramural Hematoma,IMH)以及穿透性主动脉溃疡(Penetrating Aortic Ulcer,PAU)。 核心功能: - 基于TotalSegmentator实现预处理与主动脉掩码生成 - 借助nnUNet完成AD/IMH/PAU多类别分割(需使用预训练权重) - 基于自定义.pth模型实现切片级分类 - 配备易用的图形用户界面(Graphical User Interface,GUI),用于路径配置与工作流执行 前置依赖:Python 3.x版本、医学影像库、TotalSegmentator以及指定目录下的预训练模型权重 输入输出:支持输入.nii.gz格式的CTA图像,可将结果输出至用户指定文件夹,支持灵活调整路径 AAS-DSS可实现自动化的17分区主动脉分割(基于扩展的SVS/STS分类标准),以辅助标准化的急性主动脉综合征诊断。



