Camurdan Ozden, Tanyel Toygar, Aktufan Cerekci Esma, Alis Deniz, Meltem Emine, Denizoglu Nurper, Seker Mustafa Ege, Oksuz Ilkay, Karaarslan Ercan
Department of Radiology, Acibadem Healthcare Group, Istanbul, Turkey.
Biomedical Engineering Graduate Program, Istanbul Technical University, Istanbul, Turkey.
Insights Imaging. 2025 Mar 19;16(1):60. doi: 10.1186/s13244-025-01922-w.
To develop an efficient deep learning (DL) model for breast cancer detection in mammograms, utilizing both weak (image-level) and strong (bounding boxes) annotations and providing explainable artificial intelligence (XAI) with gradient-weighted class activation mapping (Grad-CAM), assessed by the ground truth overlap ratio.
Three radiologists annotated a balanced dataset of 1976 mammograms (cancer-positive and -negative) from three centers. We developed a patch-based DL model using curriculum learning, progressively increasing patch sizes during training. The model was trained under varying levels of strong supervision (0%, 20%, 40%, and 100% of the dataset), resulting in baseline, curriculum 20, curriculum 40, and curriculum 100 models. Training for each model was repeated ten times, with results presented as mean ± standard deviation. Model performance was also tested on an external dataset of 4276 mammograms to assess generalizability.
F1 scores for the baseline, curriculum 20, curriculum 40, and curriculum 100 models were 80.55 ± 0.88, 82.41 ± 0.47, 83.03 ± 0.31, and 83.95 ± 0.55, respectively, with ground truth overlap ratios of 60.26 ± 1.91, 62.13 ± 1.2, 62.26 ± 1.52, and 64.18 ± 1.37. In the external dataset, F1 scores were 74.65 ± 1.35, 77.77 ± 0.73, 78.23 ± 1.78, and 78.73 ± 1.25, respectively, maintaining a similar performance trend.
Training DL models with a curriculum method and a patch-based approach yields satisfactory performance and XAI, even with a limited set of densely annotated data, offering a promising avenue for deploying DL in large-scale mammography datasets.
This study introduces a DL model for mammography-based breast cancer detection, utilizing curriculum learning with limited, strongly labeled data. It showcases performance gains and better explainability, addressing challenges of extensive dataset needs and DL's "black-box" nature.
Increasing numbers of mammograms for radiologists to interpret pose a logistical challenge. We trained a DL model leveraging curriculum learning with mixed annotations for mammography. The DL model outperformed the baseline model with image-level annotations using only 20% of the strong labels. The study addresses the challenge of requiring extensive datasets and strong supervision for DL efficacy. The model demonstrated improved explainability through Grad-CAM, verified by a higher ground truth overlap ratio. He proposed approach also yielded robust performance on external testing data.
利用弱（图像级）和强（边界框）注释开发一种用于乳腺钼靶片中乳腺癌检测的高效深度学习（DL）模型，并通过梯度加权类激活映射（Grad-CAM）提供可解释人工智能（XAI），通过真实重叠率进行评估。
三位放射科医生对来自三个中心的1976张乳腺钼靶片（癌症阳性和阴性）的平衡数据集进行了注释。我们使用课程学习开发了一个基于补丁的DL模型，在训练过程中逐步增加补丁大小。该模型在不同强度的监督水平（数据集的0%、20%、40%和100%）下进行训练，从而得到基线、课程20、课程40和课程100模型。每个模型的训练重复十次，结果以平均值±标准差表示。模型性能还在一个包含4276张乳腺钼靶片的外部数据集上进行了测试，以评估其通用性。
基线、课程20、课程40和课程100模型的F1分数分别为80.55±0.88、82.41±0.47、83.03±0.31和83.95±0.55，真实重叠率分别为60.26±1.91、62.13±1.2、62.26±1.52和64.18±1.37。在外部数据集中，F1分数分别为74.65±1.35、77.77±0.73、78.23±1.78和78.73±1.25，保持了相似的性能趋势。
即使只有有限的密集注释数据，采用课程方法和基于补丁的方法训练DL模型也能产生令人满意的性能和XAI，为在大规模乳腺钼靶数据集上部署DL提供了一条有前景的途径。
本研究介绍了一种基于乳腺钼靶片的乳腺癌检测DL模型，利用课程学习和有限的强标记数据。它展示了性能提升和更好的可解释性，解决了广泛数据集需求和DL“黑箱”性质的挑战。
放射科医生需要解读的乳腺钼靶片数量不断增加，这带来了后勤方面挑战。我们利用课程学习和混合注释训练了一个用于乳腺钼靶片的DL模型。该DL模型仅使用20%的强标签就优于具有图像级注释的基线模型。该研究解决了DL有效性需要大量数据集和强监督的挑战。该模型通过Grad-CAM展示了改进的可解释性，并通过更高的真实重叠率得到验证。他提出的方法在外部测试数据上也产生了稳健的性能。
