Machine learning techniques for mitoses classification.
Machine learning techniques for mitoses classification.
复制标题
有丝分裂分类的机器学习技术。
DOI：
10.1016/j.compmedimag.2020.101832
复制
发表时间：
2021-01
期刊：
影响因子：
--
通讯作者：
Shapiro L
中科院分区：
文献类型：
--
作者：
Nofallah S;Mehta S;Mercan E;Knezevich S;May CJ;Weaver D;Witten D;Elmore JG;Shapiro L
病理学家在细胞和结构水平上分析活检材料，以确定诊断和癌症分期。有丝分裂像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重要因素。卷积神经网络（CNN）在几个识别任务上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。我们研究了两种最先进的基于CNN的模型，ESPNet和DenseNet，用于皮肤活检的六个完整切片图像的有丝分裂分类，并比较了它们在灵敏度，特异性和F评分方面的定量性能。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，我们在公开的MITOS乳腺活检数据集上研究了这些分类器和其他两种架构ResNet和ShuffleNet的性能，并在精度，召回率和F分数（这是该数据集的标准），架构，训练时间和推理时间方面比较了所有四种分类器的性能。我们的原发性黑色素瘤数据集的ESPNet和DenseNet结果的灵敏度分别为0.976和0.968，特异性分别为0.987和0.995，F评分分别为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。使用DenseNet的MITOS结果的精确度为0.939，召回率为0.916，F分数为0.927。MITOS上发表的最佳结果（Saha等人）精确度为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方面击败了其他人（DenseNet 0.927，ESPNet 0.890，ResNet 0.865，ShuffleNet 0.847），尤其是Recall（DenseNet 0.916，ESPNet 0.866，ResNet 0.807，ShuffleNet 0.753），而ResNet和ESPNet的推理时间要快得多（ResNet 6秒，ESPNet 8秒，DenseNet 31秒）。ResNet比ESPNet更快，但ESPNet的F-Score和Recall比ResNet更高，使其成为一个很好的折衷解决方案。我们研究了几种最先进的CNN，用于检测整个切片活检图像中的有丝分裂像。我们在一个黑色素瘤癌症数据集上评估了两个CNN，然后在一个公共乳腺癌数据集上比较了四个CNN，两者使用相同的方法。我们在黑色素瘤和乳腺癌全切片图像中发现有丝分裂的方法和架构已经过彻底测试，并且可能对在任何全切片活检图像中发现有丝分裂有用。
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.) 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 seconds, ESPNet 8 seconds, DenseNet 31 seconds). 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.
登录
查看更多内容
影响因子：
13.8
作者：
Piepkorn, Michael W.;Barnhill, Raymond L.;Elmore, Joann G.
通讯作者：
Elmore, Joann G.
影响因子：
64.8
作者：
Esteva A;Kuprel B;Novoa RA;Ko J;Swetter SM;Blau HM;Thrun S
通讯作者：
Thrun S
DOI：
10.1001/jama.2015.1405
发表时间：
2015-03-17
期刊：
JAMA
影响因子：
--
作者：
Elmore JG;Longton GM;Carney PA;Geller BM;Onega T;Tosteson AN;Nelson HD;Pepe MS;Allison KH;Schnitt SJ;O'Malley FP;Weaver DL
通讯作者：
Weaver DL
DeepMitosis：通过深度检测、验证和分割网络进行有丝分裂检测
影响因子：
10.9
作者：
Li, Chao;Wang, Xinggang;Latecki, Longin Jan
通讯作者：
Latecki, Longin Jan
DOI：
10.1136/bmj.j2813
发表时间：
2017-06-28
期刊：
BMJ (Clinical research ed.)
影响因子：
--
作者：
Elmore JG;Barnhill RL;Elder DE;Longton GM;Pepe MS;Reisch LM;Carney PA;Titus LJ;Nelson HD;Onega T;Tosteson ANA;Weinstock MA;Knezevich SR;Piepkorn MW
通讯作者：
Piepkorn MW