文献速递眼底病与葡萄膜炎

深度学习算法与积分梯度解释辅助糖尿病视网膜病变分级

Using a Deep Learning Algorithm and Integrated Gradients Explanation to Assist Grading for Diabetic Retinopathy.
PubMed 文献 · Ophthalmology · 2018-12-13

结果显示模型辅助可提高分级准确性,尤其对中度及以上DR的敏感性从79.4%升至87.5%或88.7%且特异性未显著下降,但热力图对无DR患者降低准确性。

要点

  • 模型辅助分级可提高医生对糖尿病视网膜病变分级的准确性(P<0.001)。
  • 分级加热力图对DR患者提高准确性(P<0.001),但对无DR患者降低准确性(P=0.006)。
  • 两种辅助均提高中度及以上DR敏感性且不降低特异性,并增加分级信心和时间。

核心结论

  • 深度学习算法辅助可提高医生诊断糖尿病视网膜病变的准确性和信心,但可能增加分级时间。

主要发现

  • 仅分级辅助比无辅助分级更准确(P<0.001)。
  • 分级加热力图对DR患者提高准确性(P<0.001),对无DR患者降低准确性(P=0.006)。
  • 模型辅助提高视网膜专科医生准确性,超过无辅助读者或模型单独。
  • 模型辅助增加所有读者分级信心和分级时间。
  • 多数情况下分级加热力图仅与仅分级同样有效。

关键数值与推荐

  • 中度及以上DR敏感性:无辅助均值79.4%(95%CI 72.3%-86.5%),仅分级均值87.5%(95%CI 85.1%-89.9%),分级加热力图均值88.7%(95%CI 84.9%-92.5%)。
  • 特异性:无辅助均值96.6%(95%CI 95.9%-97.4%),仅分级均值96.1%(95%CI 95.5%-96.7%),分级加热力图均值95.5%(95%CI 94.8%-96.1%)。

临床怎么用

  • 临床中可借助深度学习模型分级辅助提高DR诊断准确性,尤其对中度及以上病变敏感性提升。
  • 热力图解释可能对无DR患者产生误导,需谨慎使用。
  • 辅助可能延长分级时间,但随经验增加时间可减少。

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

PURPOSE: To understand the impact of deep learning diabetic retinopathy (DR) algorithms on physician readers in computer-assisted settings. DESIGN: Evaluation of diagnostic technology. PARTICIPANTS: One thousand seven hundred ninety-six retinal fundus images from 1612 diabetic patients. METHODS: Ten ophthalmologists (5 general ophthalmologists, 4 retina specialists, 1 retina fellow) read images for DR severity based on the International Clinical Diabetic Retinopathy disease severity scale in each of 3 conditions: unassisted, grades only, or grades plus heatmap. Grades-only assistance comprised a histogram of DR predictions (grades) from a trained deep-learning model. For grades plus heatmap, we additionally showed explanatory heatmaps. MAIN OUTCOME MEASURES: For each experiment arm, we computed sensitivity and specificity of each reader and the algorithm for different levels of DR severity against an adjudicated reference standard. We also measured accuracy (exact 5-class level agreement and Cohen's quadratically weighted κ), reader-reported confidence (5-point Likert scale), and grading time. RESULTS: Readers graded more accurately with model assistance than without for the grades-only condition (P < 0.001). Grades plus heatmaps improved accuracy for patients with DR (P < 0.001), but reduced accuracy for patients without DR (P = 0.006). Both forms of assistance increased readers' sensitivity moderate-or-worse DR: unassisted: mean, 79.4% [95% confidence interval (CI), 72.3%-86.5%]; grades only: mean, 87.5% [95% CI, 85.1%-89.9%]; grades plus heatmap: mean, 88.7% [95% CI, 84.9%-92.5%] without a corresponding drop in specificity (unassisted: mean, 96.6% [95% CI, 95.9%-97.4%]; grades only: mean, 96.1% [95% CI, 95.5%-96.7%]; grades plus heatmap: mean, 95.5% [95% CI, 94.8%-96.1%]). Algorithmic assistance increased the accuracy of retina specialists above that of the unassisted reader or model alone; and increased grading confidence and grading time across all readers. For most cases, grades plus heatmap was only as effective as grades only. Over the course of the experiment, grading time decreased across all conditions, although most sharply for grades plus heatmap. CONCLUSIONS: Deep learning algorithms can improve the accuracy of, and confidence in, DR diagnosis in an assisted read setting. They also may increase grading time, although these effects may be ameliorated with experience.

来源:Ophthalmology,查看原文。版权归原作者所有。

引用

  1. Sayres R, Taly A, Rahimy E, et al. Using a Deep Learning Algorithm and Integrated Gradients Explanation to Assist Grading for Diabetic Retinopathy. Ophthalmology. 2018. PMID: 30553900. doi:10.1016/j.ophtha.2018.11.016

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。