Deep Learning-Based High-Frequency Ultrasound Skin Image Classification with Multicriteria Model Evaluation.

Deep Learning-Based High-Frequency Ultrasound Skin Image Classification with Multicriteria Model Evaluation.
复制标题

DOI:
10.3390/s21175846
复制
发表时间:
2021-08-30
期刊:
Sensors (Basel, Switzerland)
影响因子:
--
通讯作者:
Słowińska M
Słowińska M
中科院分区:
其他
文献类型:
--
作者:
Czajkowska J;Badura P;Korzekwa S;Płatkowska-Szczerek A;Słowińska M

文献摘要

参考文献

被引文献

相似文献

本研究首次将卷积神经网络应用于高频超声皮肤图像分类。这种类型的成像为皮肤病学开辟了新的机会,显示炎症性疾病,如特应性皮炎、牛皮癣或皮肤病变。我们收集了631张健康皮肤和不同皮肤病理图像的数据库,以训练和评估方法的各个阶段。提出的框架首先使用DeepLab v3+模型和预训练的异常主干对表皮层进行分割。我们使用迁移学习来训练分割模型有两个目的:提取感兴趣的区域进行分类,并准备用于分类置信度估计的皮肤层图。为了分类,我们在不同的输入数据模式和数据增强设置下训练了五个模型。我们还引入了分类置信水平来评估深度模型的可靠性。该测量将我们的皮肤层图与由Grad-CAM技术生成的热图相结合,该技术旨在指示深度模型用于做出分类决策的图像区域。此外,我们提出了一种多准则模型评价方法,从分类精度、置信度和测试数据集大小三个方面选择最优模型。本文的实验表明,用提取的感兴趣区域馈送DenseNet-201模型可以得到最可靠、最准确的结果。
This study presents the first application of convolutional neural networks to high-frequency ultrasound skin image classification. This type of imaging opens up new opportunities in dermatology, showing inflammatory diseases such as atopic dermatitis, psoriasis, or skin lesions. We collected a database of 631 images with healthy skin and different skin pathologies to train and assess all stages of the methodology. The proposed framework starts with the segmentation of the epidermal layer using a DeepLab v3+ model with a pre-trained Xception backbone. We employ transfer learning to train the segmentation model for two purposes: to extract the region of interest for classification and to prepare the skin layer map for classification confidence estimation. For classification, we train five models in different input data modes and data augmentation setups. We also introduce a classification confidence level to evaluate the deep model’s reliability. The measure combines our skin layer map with the heatmap produced by the Grad-CAM technique designed to indicate image regions used by the deep model to make a classification decision. Moreover, we propose a multicriteria model evaluation measure to select the optimal model in terms of classification accuracy, confidence, and test dataset size. The experiments described in the paper show that the DenseNet-201 model fed with the extracted region of interest produces the most reliable and accurate results.
DOI: 10.3390/cancers13040738
发表时间: 2021-02-10
期刊: Cancers
影响因子: 5.2
作者:
Ayana G;Dese K;Choe SW
通讯作者: Choe SW
DOI: 10.1016/j.compmedimag.2019.101676
发表时间: 2020-01-01
影响因子: 5.7
作者:
Czajkowska, Joanna;Korzekwa, Szymon;Pietka, Ewa
通讯作者: Pietka, Ewa
DOI: 10.1016/j.ultras.2021.106412
发表时间: 2021-03-27
期刊: ULTRASONICS
影响因子: 4.2
作者:
Czajkowska, Joanna;Badura, Pawel;Platkowska-Szczerek, Anna
通讯作者: Platkowska-Szczerek, Anna
DOI: 10.1016/j.ultrasmedbio.2013.04.006
发表时间: 2013-11-01
影响因子: 2.9
作者:
Gao, Yi;Tannenbaum, Allen;Liu, Tian
通讯作者: Liu, Tian
DOI: 10.15386/mpr-1658
发表时间: 2021-01
影响因子: --
作者:
Bezugly A;Sedova T;Belkov P;Enikeev D;Voloshin R
通讯作者: Voloshin R