Whata Albert, Dibeco Katlego, Madzima Kudakwashe, Obagbuwa Ibidun
Department of Mathematical Sciences, Sol Plaatje University, Kimberley, South Africa.
Department of Computer Science and Information Technology, Sol Plaatje University, Kimberley, South Africa.
Front Artif Intell. 2024 Sep 18;7:1410841. doi: 10.3389/frai.2024.1410841. eCollection 2024.
This paper investigates uncertainty quantification (UQ) techniques in multi-class classification of chest X-ray images (COVID-19, Pneumonia, and Normal). We evaluate Bayesian Neural Networks (BNN) and the Deep Neural Network with UQ (DNN with UQ) techniques, including Monte Carlo dropout, Ensemble Bayesian Neural Network (EBNN), Ensemble Monte Carlo (EMC) dropout, across different evaluation metrics. Our analysis reveals that DNN with UQ, especially EBNN and EMC dropout, consistently outperform BNNs. For example, in Class 0 vs. All, EBNN achieved a Acc of 92.6%, AUC-ROC of 95.0%, and a Brier Score of 0.157, significantly surpassing BNN's performance. Similarly, EMC Dropout excelled in Class 1 vs. All with a Acc of 83.5%, AUC-ROC of 95.8%, and a Brier Score of 0.165. These advanced models demonstrated higher accuracy, better discriaminative capability, and more accurate probabilistic predictions. Our findings highlight the efficacy of DNN with UQ in enhancing model reliability and interpretability, making them highly suitable for critical healthcare applications like chest X-ray imageQ6 classification.
本文研究胸部X光图像（新冠肺炎、肺炎和正常）多类别分类中的不确定性量化（UQ）技术。我们评估了贝叶斯神经网络（BNN）以及采用UQ技术的深度神经网络（带UQ的DNN），包括蒙特卡洛随机失活、集成贝叶斯神经网络（EBNN）、集成蒙特卡洛（EMC）随机失活，并采用了不同的评估指标。我们的分析表明，带UQ的DNN，尤其是EBNN和EMC随机失活，始终优于BNN。例如，在类别0与所有其他类别对比中，EBNN的准确率达到92.6%，AUC-ROC为95.0%，布里尔分数为0.157，显著超过了BNN的性能。同样，EMC随机失活在类别1与所有其他类别对比中表现出色，准确率为83.5%，AUC-ROC为95.8%，布里尔分数为0.165。这些先进模型展现出更高的准确性、更好的判别能力以及更准确的概率预测。我们的研究结果突出了带UQ的DNN在提高模型可靠性和可解释性方面的有效性，使其非常适用于胸部X光图像分类等关键医疗保健应用。
