Department of Stomatology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China.
College of Intelligent Transportation, Zhejiang Institute of Communications, Hangzhou, China.
J Oral Pathol Med. 2024 May;53(5):294-302. doi: 10.1111/jop.13536. Epub 2024 Apr 17.
Early diagnosis in oral cancer is essential to reduce both morbidity and mortality. This study explores the use of uncertainty estimation in deep learning for early oral cancer diagnosis.
We develop a Bayesian deep learning model termed 'Probabilistic HRNet', which utilizes the ensemble MC dropout method on HRNet. Additionally, two oral lesion datasets with distinct distributions are created. We conduct a retrospective study to assess the predictive performance and uncertainty of Probabilistic HRNet across these datasets.
Probabilistic HRNet performs optimally on the In-domain test set, achieving an F1 score of 95.3% and an AUC of 96.9% by excluding the top 30% high-uncertainty samples. For evaluations on the Domain-shift test set, the results show an F1 score of 64.9% and an AUC of 80.3%. After excluding 30% of the high-uncertainty samples, these metrics improve to an F1 score of 74.4% and an AUC of 85.6%.
Redirecting samples with high uncertainty to experts for subsequent diagnosis significantly decreases the rates of misdiagnosis, which highlights that uncertainty estimation is vital to ensure safe decision making for computer-aided early oral cancer diagnosis.
早期诊断口腔癌对于降低发病率和死亡率至关重要。本研究探讨了在深度学习中使用不确定性估计进行早期口腔癌诊断。
我们开发了一种称为“概率性 HRNet”的贝叶斯深度学习模型，该模型在 HRNet 上使用集成的 MC 辍学方法。此外，创建了两个具有不同分布的口腔病变数据集。我们进行了一项回顾性研究，以评估 Probabilistic HRNet 在这些数据集上的预测性能和不确定性。
Probabilistic HRNet 在内部测试集上表现最佳，通过排除前 30%的高不确定性样本，其 F1 得分达到 95.3%，AUC 达到 96.9%。对于域转移测试集的评估，结果显示 F1 得分为 64.9%，AUC 为 80.3%。排除 30%的高不确定性样本后，这些指标提高到 F1 得分为 74.4%，AUC 为 85.6%。
将高不确定性样本重新引导给专家进行后续诊断，可以显著降低误诊率，这表明不确定性估计对于确保计算机辅助早期口腔癌诊断的安全决策至关重要。
Zotero 插件