Song Bofan, Sunny Sumsum, Li Shaobai, Gurushanth Keerthi, Mendonca Pramila, Mukhia Nirza, Patrick Sanjana, Gurudath Shubha, Raghavan Subhashini, Tsusennaro Imchen, Leivon Shirley T, Kolur Trupti, Shetty Vivek, Bushan Vidya R, Ramesh Rohan, Peterson Tyler, Pillai Vijay, Wilder-Smith Petra, Sigamani Alben, Suresh Amritha, Kuriakose Moni Abraham, Birur Praveen, Liang Rongguang
Wyant College of Optical Sciences, The University of Arizona, Tucson, Arizona 85721, USA.
Mazumdar Shaw Medical Centre, Bangalore, India.
Biomed Opt Express. 2021 Sep 20;12(10):6422-6430. doi: 10.1364/BOE.432365. eCollection 2021 Oct 1.
In medical imaging, deep learning-based solutions have achieved state-of-the-art performance. However, reliability restricts the integration of deep learning into practical medical workflows since conventional deep learning frameworks cannot quantitatively assess model uncertainty. In this work, we propose to address this shortcoming by utilizing a Bayesian deep network capable of estimating uncertainty to assess oral cancer image classification reliability. We evaluate the model using a large intraoral cheek mucosa image dataset captured using our customized device from high-risk population to show that meaningful uncertainty information can be produced. In addition, our experiments show improved accuracy by uncertainty-informed referral. The accuracy of retained data reaches roughly 90% when referring either 10% of all cases or referring cases whose uncertainty value is greater than 0.3. The performance can be further improved by referring more patients. The experiments show the model is capable of identifying difficult cases needing further inspection.
在医学成像中，基于深度学习的解决方案已取得了领先的性能。然而，可靠性限制了深度学习在实际医疗工作流程中的集成，因为传统的深度学习框架无法定量评估模型的不确定性。在这项工作中，我们建议通过利用能够估计不确定性的贝叶斯深度网络来解决这一缺点，以评估口腔癌图像分类的可靠性。我们使用通过定制设备从高危人群中采集的大型口腔颊黏膜图像数据集对模型进行评估，结果表明可以生成有意义的不确定性信息。此外，我们的实验表明，通过基于不确定性的转诊提高了准确率。当转诊所有病例的10%或转诊不确定性值大于0.3的病例时，保留数据的准确率大致达到90%。转诊更多患者可进一步提高性能。实验表明，该模型能够识别需要进一步检查的疑难病例。
