University of Washington, Seattle WA 98195, USA.
Pathology Associates, Clovis, CA 983611, USA.
Comput Med Imaging Graph. 2021 Jan;87:101832. doi: 10.1016/j.compmedimag.2020.101832. Epub 2020 Nov 27.
Pathologists analyze biopsy material at both the cellular and structural level to determine diagnosis and cancer stage. Mitotic figures are surrogate biomarkers of cellular proliferation that can provide prognostic information; thus, their precise detection is an important factor for clinical care. Convolutional Neural Networks (CNNs) have shown remarkable performance on several recognition tasks. Utilizing CNNs for mitosis classification may aid pathologists to improve the detection accuracy.
We studied two state-of-the-art CNN-based models, ESPNet and DenseNet, for mitosis classification on six whole slide images of skin biopsies and compared their quantitative performance in terms of sensitivity, specificity, and F-score. We used raw RGB images of mitosis and non-mitosis samples with their corresponding labels as training input. In order to compare with other work, we studied the performance of these classifiers and two other architectures, ResNet and ShuffleNet, on the publicly available MITOS breast biopsy dataset and compared the performance of all four in terms of precision, recall, and F-score (which are standard for this data set), architecture, training time and inference time.
The ESPNet and DenseNet results on our primary melanoma dataset had a sensitivity of 0.976 and 0.968, and a specificity of 0.987 and 0.995, respectively, with F-scores of .968 and .976, respectively. On the MITOS dataset, ESPNet and DenseNet showed a sensitivity of 0.866 and 0.916, and a specificity of 0.973 and 0.980, respectively. The MITOS results using DenseNet had a precision of 0.939, recall of 0.916, and F-score of 0.927. The best published result on MITOS (Saha et al. 2018) reported precision of 0.92, recall of 0.88, and F-score of 0.90. In our architecture comparisons on MITOS, we found that DenseNet beats the others in terms of F-Score (DenseNet 0.927, ESPNet 0.890, ResNet 0.865, ShuffleNet 0.847) and especially Recall (DenseNet 0.916, ESPNet 0.866, ResNet 0.807, ShuffleNet 0.753), while ResNet and ESPNet have much faster inference times (ResNet 6 s, ESPNet 8 s, DenseNet 31 s). ResNet is faster than ESPNet, but ESPNet has a higher F-Score and Recall than ResNet, making it a good compromise solution.
We studied several state-of-the-art CNNs for detecting mitotic figures in whole slide biopsy images. We evaluated two CNNs on a melanoma cancer dataset and then compared four CNNs on a public breast cancer data set, using the same methodology on both. Our methodology and architecture for mitosis finding in both melanoma and breast cancer whole slide images has been thoroughly tested and is likely to be useful for finding mitoses in any whole slide biopsy images.
病理学家通过细胞和结构水平分析活检材料来确定诊断和癌症分期。有丝分裂图是细胞增殖的替代生物标志物，可以提供预后信息；因此，它们的精确检测是临床护理的一个重要因素。卷积神经网络 (CNN) 在多个识别任务中表现出了卓越的性能。利用 CNN 进行有丝分裂分类可以帮助病理学家提高检测精度。
我们研究了两种基于最先进的 CNN 的模型，ESPNet 和 DenseNet，用于对六张皮肤活检全幻灯片图像中的有丝分裂进行分类，并根据灵敏度、特异性和 F 分数比较它们的定量性能。我们使用有丝分裂和非有丝分裂样本的原始 RGB 图像及其相应的标签作为训练输入。为了与其他工作进行比较，我们研究了这些分类器和另外两个架构（ResNet 和 ShuffleNet）在公开可用的 MITOS 乳腺活检数据集上的性能，并根据精度、召回率和 F 分数（这是该数据集的标准）比较了所有四个架构的性能，架构，训练时间和推理时间。
ESPNet 和 DenseNet 在我们的主要黑色素瘤数据集上的结果具有 0.976 和 0.968 的灵敏度，0.987 和 0.995 的特异性，F 分数分别为 0.968 和 0.976。在 MITOS 数据集上，ESPNet 和 DenseNet 的灵敏度分别为 0.866 和 0.916，特异性分别为 0.973 和 0.980。DenseNet 在 MITOS 上的结果具有 0.939 的精度、0.916 的召回率和 0.927 的 F 分数。MITOS 上发表的最佳结果（Saha 等人，2018 年）报告的精度为 0.92、召回率为 0.88、F 分数为 0.90。在我们对 MITOS 的架构比较中，我们发现 DenseNet 在 F-Score（DenseNet 0.927、ESPNet 0.890、ResNet 0.865、ShuffleNet 0.847）方面优于其他架构，尤其是在召回率方面（DenseNet 0.916、ESPNet 0.866、ResNet 0.807、ShuffleNet 0.753），而 ResNet 和 ESPNet 的推理时间更快（ResNet 6 秒，ESPNet 8 秒，DenseNet 31 秒）。ResNet 比 ESPNet 快，但 ESPNet 的 F-Score 和召回率都高于 ResNet，因此是一个很好的折衷方案。
我们研究了几种用于检测全幻灯片活检图像中有丝分裂图的最先进的 CNN。我们在黑色素瘤癌症数据集上评估了两个 CNN，然后在公共乳腺癌数据集上比较了四个 CNN，在两个数据集上使用相同的方法。我们在黑色素瘤和乳腺癌全幻灯片图像中进行有丝分裂检测的方法和架构已经过全面测试，可能对任何全幻灯片活检图像中的有丝分裂检测都有用。
