BACKGROUND:Since Gleason score (GS) 4 + 3 prostate cancer (PCa) has the worse prognosis than GS 3 + 4 PCa, differentiating these two types of PCa is of clinical significance. OBJECTIVE:To assess the predictive roles of using T2WI and ADC-derived image texture parameters in differentiating GS 3 + 4from GS 4 + 3 PCa. METHODS:Forty-eight PCa patients of GS 3 + 4 and 37 patients of GS 4 + 3 are retrieved and randomly divided into training (60%) and testing (40%) sets. Axial image showing the maximum tumor size is selected in the T2WI and ADC maps for further image texture feature analysis. Three hundred texture features are computed from each region of interest (ROI) using MaZda software. Feature reduction is implemented to obtain 30 optimal features, which are then used to generate the most discriminative features (MDF). Receiver operating characteristic (ROC) curve analysis is performed on MDF values in the training sets to achieve cutoff values for determining the correct rates of discrimination between two Gleason patterns in the testing sets. RESULTS:ROC analysis on T2WI and ADC-derived MDF values in the training set (n = 51) results in a mean area under the curve (AUC) of 0.953±0.025 (with sensitivity 92.74±6.15 and specificity 89.7±6.9), and 0.985±0.013 (with sensitivity 96.36±4.46 and specificity 97.26±2.58), respectively. Using the corresponding MDF cutoffs, 95.3% (ranges from 76.5% to 100%) and 94.1% (ranged from 76.5% to 100%) of test cases (n = 34) are correctly discriminated using T2WI and ADC-derived MDF values, respectively. CONCLUSIONS:The study demonstrates that using T2WI and ADC-derived image texture parameters has a potentially predictive role in differentiating GS 3 + 4 and GS 4 + 3 PCa.

中文翻译：

---

T2WI 和 ADC 衍生的纹理参数在区分 Gleason 评分 3 + 4 和 4 + 3 前列腺癌中的预测作用

背景：由于Gleason评分（GS）4+3前列腺癌（PCa）的预后比GS 3+4PCa差，因此区分这两种类型的PCa具有临床意义。目的：评估使用 T2WI 和 ADC 衍生的图像纹理参数在区分 GS 3 + 4 和 GS 4 + 3 PCa 中的预测作用。方法：检索 48 例 GS 3+4 的 PCa 患者和 37 例 GS 4+3 的患者，随机分为训练集（60%）和测试集（40%）。在 T2WI 和 ADC 图中选择显示最大肿瘤大小的轴位图像以进行进一步的图像纹理特征分析。使用 MaZda 软件从每个感兴趣区域 (ROI) 计算出三百个纹理特征。执行特征缩减以获得 30 个最佳特征，然后将其用于生成最具辨别力的特征 (MDF)。对训练集中的 MDF 值进行接受者操作特征 (ROC) 曲线分析，以获得确定测试集中两个格里森模式之间正确区分率的截止值。结果：训练集（n = 51）中 T2WI 和 ADC 衍生的 MDF 值的 ROC 分析导致平均曲线下面积 (AUC) 为 0.953±0.025（灵敏度为 92.74±6.15，特异性为 89.7±6.9），以及分别为 0.985±0.013（灵敏度 96.36±4.46 和特异性 97.26±2.58）。使用相应的 MDF 临界值，分别使用 T2WI 和 ADC 导出的 MDF 值正确区分了 95.3%（范围从 76.5% 到 100%）和 94.1%（范围从 76.5% 到 100%）的测试用例（n = 34） . 结论：