Harbin Institute of Technology, Harbin, 150001, Heilongjiang China.
Heilongjiang Provincial Hospital, Harbin, 150001, Heilongjiang, China.
Comput Methods Programs Biomed. 2022 Jun;221:106911. doi: 10.1016/j.cmpb.2022.106911. Epub 2022 May 25.
Grading the severity level is an extremely important procedure for correct diagnoses and personalized treatment schemes for acne. However, the acne grading criteria are not unified in the medical field. This work aims to develop an acne diagnosis system that can be generalized to various criteria.
A unified acne grading framework that can be generalized to apply referring to different grading criteria is developed. It imitates the global estimation of the dermatologist diagnosis in two steps. First, an adaptive image preprocessing method effectively filters meaningless information and enhances key information. Next, an innovative network structure fuses global deep features with local features to simulate the dermatologists' comparison of local skin and global observation. In addition, a transfer fine-tuning strategy is proposed to transfer prior knowledge on one criterion to another criterion, which effectively improves the framework performance in case of insufficient data.
The Preprocessing method effectively filters meaningless areas and improves the performance of downstream models.The framework reaches accuracies of 84.52% and 59.35% on two datasets separately.
The application of the framework on acne grading exceeds the state-of-the-art method by 1.71%, reaches the diagnostic level of a professional dermatologist and the transfer fine-tuning strategy improves the accuracy of 6.5% on the small data.
对痤疮严重程度进行分级是正确诊断和制定个体化治疗方案的关键步骤。然而，目前医学领域的痤疮分级标准尚未统一。本研究旨在开发一种适用于多种分级标准的痤疮诊断系统。
开发了一种可以推广到不同分级标准的统一痤疮分级框架，它模拟了皮肤科医生诊断的两步全局评估。首先，采用自适应图像预处理方法有效滤除无意义信息并增强关键信息。其次，创新的网络结构融合全局深度特征和局部特征，模拟皮肤科医生对局部皮肤和整体观察的比较。此外，提出了一种迁移微调策略，将一个标准的先验知识转移到另一个标准，在数据不足的情况下有效提高了框架的性能。
预处理方法有效滤除无意义区域，提高了下游模型的性能。该框架在两个数据集上的准确率分别达到 84.52%和 59.35%。
该框架在痤疮分级中的应用比现有方法提高了 1.71%，达到了专业皮肤科医生的诊断水平，并且迁移微调策略可将小数据集的准确率提高 6.5%。
