Wang Hu, Butler David, Zhang Yuan, Avery Jodie, Knox Steven, Ma Congbo, Hull Louise, Carneiro Gustavo
The University of Adelaide, Adelaide, Australia.
Benson Radiology, Adelaide, Australia.
Phys Med Biol. 2024 Dec 24;70(1). doi: 10.1088/1361-6560/ad997e.
Endometriosis, affecting about 10% of individuals assigned female at birth, is challenging to diagnose and manage. Diagnosis typically involves the identification of various signs of the disease using either laparoscopic surgery or the analysis of T1/T2 MRI images, with the latter being quicker and cheaper but less accurate. A key diagnostic sign of endometriosis is the obliteration of the pouch of Douglas (POD). However, even experienced clinicians struggle with accurately classifying POD obliteration from MRI images, which complicates the training of reliable AI models.In this paper, we introduce theuman-llaborativeulti-modalulti-rater Learning (HAICOMM) methodology to address the challenge above. HAICOMM is the first method that explores three important aspects of this problem: (1) multi-rater learning to extract a cleaner label from the multiple 'noisy' labels available per training sample; (2) multi-modal learning to leverage the presence of T1/T2 MRI images for training and testing; and (3) human-AI collaboration to build a system that leverages the predictions from clinicians and the AI model to provide more accurate classification than standalone clinicians and AI models.Presenting results on the multi-rater T1/T2 MRI endometriosis dataset collected for validating the methodology, the proposed HAICOMM model outperforms an ensemble of clinicians, noisy-label learning models, and multi-rater learning methods by a large margin.The HAICOMM methodology offers a novel solution to the long-standing problem of accurately diagnosing endometriosis from MRI images, specifically in relation to the key diagnostic sign of POD obliteration. By leveraging multi-rater, multi-modal, and human-AI collaborative learning, it has the potential to improve the accuracy of endometriosis diagnosis, which could have far-reaching implications for the better management of this challenging medical condition that affects a significant proportion of the female population.
子宫内膜异位症影响着约10%出生时被指定为女性的个体，其诊断和管理具有挑战性。诊断通常涉及通过腹腔镜手术或T1/T2 MRI图像分析来识别该疾病的各种体征，后者更快、更便宜，但准确性较低。子宫内膜异位症的一个关键诊断体征是道格拉斯窝（POD）闭塞。然而，即使是经验丰富的临床医生也难以从MRI图像中准确分类POD闭塞，这使得可靠的人工智能模型的训练变得复杂。在本文中，我们引入了人机协作多模态多评分者学习（HAICOMM）方法来应对上述挑战。HAICOMM是第一种探索该问题三个重要方面的方法：（1）多评分者学习，从每个训练样本可用的多个“噪声”标签中提取更清晰的标签；（2）多模态学习，利用T1/T2 MRI图像进行训练和测试；（3）人机协作，构建一个利用临床医生和人工智能模型的预测来提供比单独的临床医生和人工智能模型更准确分类的系统。在为验证该方法而收集的多评分者T1/T2 MRI子宫内膜异位症数据集中呈现的结果表明，所提出的HAICOMM模型大大优于临床医生、噪声标签学习模型和多评分者学习方法的集合。HAICOMM方法为从MRI图像中准确诊断子宫内膜异位症这一长期存在的问题提供了一种新颖的解决方案，特别是与POD闭塞的关键诊断体征相关。通过利用多评分者、多模态和人机协作学习，它有可能提高子宫内膜异位症诊断的准确性，这可能对更好地管理这一影响相当一部分女性人群的具有挑战性的医疗状况产生深远影响。
