Vidotti Vanessa G, Costa Vital P, Silva Fabrício R, Resende Graziela M, Cremasco Fernanda, Dias Marcelo, Gomi Edson S
Glaucoma Service, Department of Ophthalmology, University of Campinas, Campinas - Brazil.
Eur J Ophthalmol. 2012 Jun 15:0. doi: 10.5301/ejo.5000183.
Purpose. To investigate the sensitivity and specificity of machine learning classifiers (MLC) and spectral domain optical coherence tomography (SD-OCT) for the diagnosis of glaucoma. Methods. Sixty-two patients with early to moderate glaucomatous visual field damage and 48 healthy individuals were included. All subjects underwent a complete ophthalmologic examination, achromatic standard automated perimetry, and RNFL imaging with SD-OCT (Cirrus HD-OCT; Carl Zeiss Meditec, Inc., Dublin, California, USA). Receiver operating characteristic (ROC) curves were obtained for all SD-OCT parameters. Subsequently, the following MLCs were tested: Classification Tree (CTREE), Random Forest (RAN), Bagging (BAG), AdaBoost M1 (ADA), Ensemble Selection (ENS), Multilayer Perceptron (MLP), Radial Basis Function (RBF), Naive-Bayes (NB), and Support Vector Machine (SVM). Areas under the ROC curves (aROCs) obtained for each parameter and each MLC were compared. Results. The mean age was 57.0±9.2 years for healthy individuals and 59.9±9.0 years for glaucoma patients (p=0.103). Mean deviation values were -4.1±2.4 dB for glaucoma patients and -1.5±1.6 dB for healthy individuals (p<0.001). The SD-OCT parameters with the greater aROCs were inferior quadrant (0.813), average thickness (0.807), 7 o'clock position (0.765), and 6 o'clock position (0.754). The aROCs from classifiers varied from 0.785 (ADA) to 0.818 (BAG). The aROC obtained with BAG was not significantly different from the aROC obtained with the best single SD-OCT parameter (p=0.93). Conclusions. The SD-OCT showed good diagnostic accuracy in a group of patients with early glaucoma. In this series, MLCs did not improve the sensitivity and specificity of SD-OCT for the diagnosis of glaucoma.
目的。研究机器学习分类器（MLC）和频域光学相干断层扫描（SD-OCT）在青光眼诊断中的敏感性和特异性。方法。纳入62例早期至中度青光眼性视野损害患者和48例健康个体。所有受试者均接受了全面的眼科检查、无色标准自动视野计检查以及使用SD-OCT（Cirrus HD-OCT；卡尔蔡司医疗技术公司，美国加利福尼亚州都柏林）进行的视网膜神经纤维层成像。获取所有SD-OCT参数的受试者工作特征（ROC）曲线。随后，测试了以下MLC：分类树（CTREE）、随机森林（RAN）、装袋法（BAG）、自适应增强M1（ADA）、集成选择（ENS）、多层感知器（MLP）、径向基函数（RBF）、朴素贝叶斯（NB）和支持向量机（SVM）。比较每个参数和每个MLC获得的ROC曲线下面积（aROC）。结果。健康个体的平均年龄为57.0±9.2岁，青光眼患者为59.9±9.0岁（p = 0.103）。青光眼患者的平均偏差值为-4.1±2.4 dB，健康个体为-1.5±1.6 dB（p<0.001）。aROC较大的SD-OCT参数为下象限（0.813）、平均厚度（0.807）、7点钟位置（0.765）和6点钟位置（0.754）。分类器的aROC范围为0.785（ADA）至0.818（BAG）。BAG获得的aROC与最佳单一SD-OCT参数获得的aROC无显著差异（p = 0.93）。结论。SD-OCT在一组早期青光眼患者中显示出良好的诊断准确性。在本系列研究中，MLC并未提高SD-OCT对青光眼诊断的敏感性和特异性。
