文献速递细胞病理与分子病理

智能手机图像结合机器学习用于甲状腺细胞病理学癌症诊断

Thyroid Cytopathology Cancer Diagnosis from Smartphone Images Using Machine Learning.
PubMed 文献 · Mod Pathol · 2023-02-13

本研究比较深度学习模型在甲状腺细针穿刺(FNA)活检分类中的表现,图像分别来自高分辨率扫描仪和手机相机。训练集为964张高分辨率扫描全切片图像,测试集为100张切片、每张手动选取20个感兴趣区(ROI)并以两种方式采集。基线模型用于手机ROI时性能下降(AUC 97.8% vs 89.5%,P=.019),训练中加入颜色增强可缩小差距(97.6% vs 96.0%,P=.309),且与病理医生表现相当。

要点

  • 基线模型对手机图像颜色变化敏感,跨模态应用时AUC由97.8%降至89.5%。
  • 训练时加入颜色增强后手机图像AUC达96.0%,与扫描图像及病理医生相当。
  • 模型在约15个ROI后性能趋于平台,提示ROI采集负担可控。

核心结论

  • 训练中加入颜色增强可缩小手机与扫描仪图像间的性能差距,使手机图像分类性能与病理医生相当。

研究设计与样本

  • 训练集为964张高分辨率扫描全切片图像;测试集为100张切片,每张手动选取20个ROI并以扫描仪和手机两种方式采集。
  • 以受试者工作特征曲线下面积(AUC)及95%CI比较基线与颜色增强模型的性能,并与病理医生对照。

主要发现

  • 基线模型:扫描图像AUC 97.8%(95%CI 95.4%–100.0%),手机图像AUC 89.5%(95%CI 82.3%–96.6%),P=.019。
  • 颜色增强后:扫描图像AUC 97.6%(95%CI 95.0%–100.0%),手机图像AUC 96.0%(95%CI 91.8%–100.0%),P=.309。
  • 病理医生恶性预测AUC 95.6%(95%CI 91.6%–99.5%),与扫描图像比较P=.398,与手机图像比较P=.875。
  • 不确定病例(Bethesda 3、4、5类):基线模型AUC 88.3%(95%CI 73.7%–100.0%),颜色增强后96.2%(95%CI 90.9%–100.0%),P=.158。
  • 模型性能在15个ROI后趋于平台;预测TBS 2的恶性率为0%,预测TBS 6的恶性率为100%。

临床怎么用

  • 提示手机拍摄的甲状腺FNA显微图像经颜色增强训练的深度学习模型可用于辅助恶性预测,尤其有望改善不确定病例的分流。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

We examined the performance of deep learning models on the classification of thyroid fine-needle aspiration biopsies using microscope images captured in 2 ways: with a high-resolution scanner and with a mobile phone camera. Our training set consisted of images from 964 whole-slide images captured with a high-resolution scanner. Our test set consisted of 100 slides; 20 manually selected regions of interest (ROIs) from each slide were captured in 2 ways as mentioned above. Applying a baseline machine learning algorithm trained on scanner ROIs resulted in performance deterioration when applied to the smartphone ROIs (97.8% area under the receiver operating characteristic curve [AUC], CI = [95.4%, 100.0%] for scanner images vs 89.5% AUC, CI = [82.3%, 96.6%] for mobile images, P = .019). Preliminary analysis via histogram matching showed that the baseline model was overly sensitive to slight color variations in the images (specifically, to color differences between mobile and scanner images). Adding color augmentation during training reduces this sensitivity and narrows the performance gap between mobile and scanner images (97.6% AUC, CI = [95.0%, 100.0%] for scanner images vs 96.0% AUC, CI = [91.8%, 100.0%] for mobile images, P = .309), with both modalities on par with human pathologist performance (95.6% AUC, CI = [91.6%, 99.5%]) for malignancy prediction (P = .398 for pathologist vs scanner and P = .875 for pathologist vs mobile). For indeterminate cases (pathologist-assigned Bethesda category of 3, 4, or 5), color augmentations confer some improvement (88.3% AUC, CI = [73.7%, 100.0%] for the baseline model vs 96.2% AUC, CI = [90.9%, 100.0%] with color augmentations, P = .158). In addition, we found that our model's performance levels off after 15 ROIs, a promising indication that ROI data collection would not be time-consuming for our diagnostic system. Finally, we showed that the model has sensible Bethesda category (TBS) predictions (increasing risk malignancy rate with predicted TBS category, with 0% malignancy for predicted TBS 2 and 100% malignancy for TBS 6).

来源:Mod Pathol,查看原文。版权归原作者所有。

引用

  1. Assaad S, Dov D, Davis R, et al. Thyroid Cytopathology Cancer Diagnosis from Smartphone Images Using Machine Learning. Mod Pathol. 2023. PMID: 36931041. doi:10.1016/j.modpat.2023.100129

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。