标题： 肿瘤位置加权MRI报告对比学习：提高儿童脑肿瘤诊断可解释性的框架 显示英文标题

标题： Tumor Location-weighted MRI-Report Contrastive Learning: A Framework for Improving the Explainability of Pediatric Brain Tumor Diagnosis

摘要： 尽管卷积神经网络（CNNs）在基于磁共振成像（MRI）的脑肿瘤诊断中表现出色，但它们尚未被广泛整合到临床工作流程中。这主要是因为模型预测所依赖的特征对放射科医生来说并不明确，因此在临床上被认为无关紧要，即缺乏可解释性。作为放射科医生知识和经验的宝贵来源，放射学报告可以与MRI结合在一个对比学习（CL）框架中，通过从图像-报告关联中学习来提高CNN的可解释性。 在这项工作中，我们在三维脑部MRI扫描和放射学报告上训练了一个多模态对比学习架构，以学习有信息量的MRI表示。此外，我们将与多种脑肿瘤分析任务相关的肿瘤位置引入该框架，以提高其泛化能力。然后，我们应用学习到的图像表示来改进儿童低级别胶质瘤（最常见的儿童脑肿瘤）的遗传标志物分类的可解释性和性能，作为下游任务。 我们的结果显示，模型注意力图与手动肿瘤分割之间的Dice分数为31.1%（作为可解释性的衡量标准），测试分类性能为87.7%，显著优于基线。这些改进可以在放射科医生中建立对模型的信任，促进其整合到临床实践中，以实现更高效的肿瘤诊断。 显示英文摘要

摘要： Despite the promising performance of convolutional neural networks (CNNs) in brain tumor diagnosis from magnetic resonance imaging (MRI), their integration into the clinical workflow has been limited. That is mainly due to the fact that the features contributing to a model's prediction are unclear to radiologists and hence, clinically irrelevant, i.e., lack of explainability. As the invaluable sources of radiologists' knowledge and expertise, radiology reports can be integrated with MRI in a contrastive learning (CL) framework, enabling learning from image-report associations, to improve CNN explainability. In this work, we train a multimodal CL architecture on 3D brain MRI scans and radiology reports to learn informative MRI representations. Furthermore, we integrate tumor location, salient to several brain tumor analysis tasks, into this framework to improve its generalizability. We then apply the learnt image representations to improve explainability and performance of genetic marker classification of pediatric Low-grade Glioma, the most prevalent brain tumor in children, as a downstream task. Our results indicate a Dice score of 31.1% between the model's attention maps and manual tumor segmentation (as an explainability measure) with test classification performance of 87.7%, significantly outperforming the baselines. These enhancements can build trust in our model among radiologists, facilitating its integration into clinical practices for more efficient tumor diagnosis.