用深度学习自动识别并修复心脏CT中的金属伪影，提升放疗精度。
Small metal artifact detection and inpainting in cardiac CT images
- 基于2D U-Net检测金属伪影，3D模型重建缺失结构。
- 伪影区域结构相似度达0.988，心腔分割Dice值提升至0.964。
- 适用于心律失常放疗患者，帮助恢复被金属干扰的解剖信息。
背景：植入式心脏除颤器（ICD）金属导线导致心脏CT图像出现伪影，影响术前心脏运动量化，阻碍立体定向心律失常放疗。需在已重建的CT上准确减少金属伪影，以恢复丢失的解剖信息。目的：开发一种自动检测心脏CT中金属伪影并进行图像修补的方法。方法：收集12名接受室性心动过速放射治疗患者的ECG门控4DCT扫描数据，人工勾画伪影区域。训练2D U-Net模型用于伪影分割；通过将真实患者图像的金属伪影叠加到无伪影CT上构建合成数据集，并训练3D图像修补模型以填补伪影区域。在合成数据集上评估修补后心腔自动分割性能，同时对原始患者数据进行视觉检查。结果：伪影检测模型的Dice分数为0.958 ± 0.008；修补模型的结构相似性指数为0.988 ± 0.012。心腔分割表面Dice分数从0.684 ± 0.247提升至0.964 ± 0.067，豪斯多夫距离由3.4 ± 3.9 mm降至0.7 ± 0.7 mm。视觉评估显示修补后图像具有高度合理性。结论：成功构建了两个深度学习模型，可有效检测并修补心脏CT中的金属伪影。
原文摘要 · Abstract (English)
Background: Quantification of cardiac motion on pre-treatment CT imaging for stereotactic arrhythmia radiotherapy patients is difficult due to the presence of image artifacts caused by metal leads of implantable cardioverter-defibrillators (ICDs). New methods are needed to accurately reduce the metal artifacts in already reconstructed CTs to recover the otherwise lost anatomical information. Purpose: To develop a methodology to automatically detect metal artifacts in cardiac CT scans and inpaint the affected volume with anatomically consistent structures and values. Methods: ECG-gated 4DCT scans of 12 patients who underwent cardiac radiation therapy for treating ventricular tachycardia were collected. The metal artifacts in the images were manually contoured. A 2D U-Net deep learning (DL) model was developed to segment the metal artifacts. A dataset of synthetic CTs was prepared by adding metal artifacts from the patient images to artifact-free CTs. A 3D image inpainting DL model was trained to refill the metal artifact portion in the synthetic images with realistic values. The inpainting model was evaluated by analyzing the automated segmentation results of the four heart chambers on the synthetic dataset. Additionally, the raw cardiac patient cases were qualitatively inspected. Results: The artifact detection model produced a Dice score of 0.958 +- 0.008. The inpainting model was able to recreate images with a structural similarity index of 0.988 +- 0.012. With the chamber segmentations improved surface Dice scores from 0.684 +- 0.247 to 0.964 +- 0.067 and the Hausdorff distance reduced from 3.4 +- 3.9 mm to 0.7 +- 0.7 mm. The inpainting model's use on cardiac patient CTs was visually inspected and the artifact-inpainted images were visually plausible. Conclusion: We successfully developed two deep models to detect and inpaint metal artifacts in cardiac CT images.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。