The University of Texas at Austin, Department of Biomedical Engineering, Austin, Texas, United States.
The University of Texas at Austin, Division of Dermatology, Dell Medical School, Austin, Texas, United States.
J Biomed Opt. 2022 Jun;27(6). doi: 10.1117/1.JBO.27.6.065004.
Raman spectroscopy (RS) provides an automated approach for assisting Mohs micrographic surgery for skin cancer diagnosis; however, the specificity of RS is limited by the high spectral similarity between tumors and normal tissues structures. Reflectance confocal microscopy (RCM) provides morphological and cytological details by which many features of epidermis and hair follicles can be readily identified. Combining RS with deep-learning-aided RCM has the potential to improve the diagnostic accuracy of RS in an automated fashion, without requiring additional input from the clinician.
The aim of this study is to improve the specificity of RS for detecting basal cell carcinoma (BCC) using an artificial neural network trained on RCM images to identify false positive normal skin structures (hair follicles and epidermis).
Our approach was to build a two-step classification model. In the first step, a Raman biophysical model that was used in prior work classified BCC tumors from normal tissue structures with high sensitivity. In the second step, 191 RCM images were collected from the same site as the Raman data and served as inputs for two ResNet50 networks. The networks selected the hair structure and epidermis images, respectively, within all images corresponding to the positive predictions of the Raman biophysical model with high specificity. The specificity of the BCC biophysical model was improved by moving the Raman spectra corresponding to these selected images from false positive to true negative.
Deep-learning trained on RCM images removed 52% of false positive predictions from the Raman biophysical model result while maintaining a sensitivity of 100%. The specificity was improved from 84.2% using Raman spectra alone to 92.4% by integrating Raman spectra with RCM images.
Combining RS with deep-learning-aided RCM imaging is a promising tool for guiding tumor resection surgery.
拉曼光谱（RS）提供了一种自动化的方法，可辅助用于皮肤癌诊断的 Mohs 显微外科手术；然而，由于肿瘤和正常组织结构之间存在很高的光谱相似性，RS 的特异性受到限制。反射共焦显微镜（RCM）通过提供形态学和细胞学细节，可以很容易地识别许多表皮和毛囊的特征。将 RS 与基于深度学习的 RCM 相结合，有可能以自动化的方式提高 RS 的诊断准确性，而无需临床医生提供额外的输入。
本研究的目的是通过基于 RCM 图像的人工神经网络来识别假阳性正常皮肤结构（毛囊和表皮），从而提高 RS 检测基底细胞癌（BCC）的特异性。
我们的方法是构建一个两步分类模型。在第一步中，使用在先前工作中使用的拉曼生物物理模型，从正常组织结构中以高灵敏度分类 BCC 肿瘤。在第二步中，从与拉曼数据相同的部位收集了 191 张 RCM 图像，并作为两个 ResNet50 网络的输入。这些网络分别选择毛发结构和表皮图像，对于拉曼生物物理模型的阳性预测，其特异性很高。通过将这些选定图像的拉曼光谱从假阳性转移到真阴性，提高了 BCC 生物物理模型的特异性。
基于 RCM 图像的深度学习训练从拉曼生物物理模型的结果中删除了 52%的假阳性预测，同时保持了 100%的敏感性。通过将拉曼光谱与 RCM 图像相结合，特异性从单独使用拉曼光谱的 84.2%提高到 92.4%。
将 RS 与基于深度学习的 RCM 成像相结合，是一种很有前途的工具，可用于指导肿瘤切除术。
