Department of Biomedical Engineering, Rutgers University, Piscataway, New Jersey, USA.
J Magn Reson Imaging. 2012 Jul;36(1):213-24. doi: 10.1002/jmri.23618. Epub 2012 Feb 15.
To identify and evaluate textural quantitative imaging signatures (QISes) for tumors occurring within the central gland (CG) and peripheral zone (PZ) of the prostate, respectively, as seen on in vivo 3 Tesla (T) endorectal T2-weighted (T2w) MRI.
This study used 22 preoperative prostate MRI data sets (16 PZ, 6 CG) acquired from men with confirmed prostate cancer (CaP) and scheduled for radical prostatectomy (RP). The prostate region-of-interest (ROI) was automatically delineated on T2w MRI, following which it was corrected for intensity-based acquisition artifacts. An expert pathologist manually delineated the dominant tumor regions on ex vivo sectioned and stained RP specimens as well as identified each of the studies as either a CG or PZ CaP. A nonlinear registration scheme was used to spatially align and then map CaP extent from the ex vivo RP sections onto the corresponding MRI slices. A total of 110 texture features were then extracted on a per-voxel basis from all T2w MRI data sets. An information theoretic feature selection procedure was then applied to identify QISes comprising T2w MRI textural features specific to CG and PZ CaP, respectively. The QISes for CG and PZ CaP were evaluated by means of Quadratic Discriminant Analysis (QDA) on a per-voxel basis against the ground truth for CaP on T2w MRI, mapped from corresponding histology.
The QDA classifier yielded an area under the Receiver Operating characteristic curve of 0.86 for the CG CaP studies, and 0.73 for the PZ CaP studies over 25 runs of randomized three-fold cross-validation. By comparison, the accuracy of the QDA classifier was significantly lower when (a) using all 110 texture features (with no feature selection applied), as well as (b) a randomly selected combination of texture features.
CG and PZ prostate cancers have significantly differing textural quantitative imaging signatures on T2w endorectal in vivo MRI.
分别识别和评估在体内 3 特斯拉（T）直肠内 T2 加权（T2w）MRI 上观察到的前列腺中央腺（CG）和外周区（PZ）内肿瘤的纹理定量成像特征（QIS）。
本研究使用了 22 例术前前列腺 MRI 数据集（16 例 PZ，6 例 CG），这些数据集来自经证实患有前列腺癌（CaP）并计划接受根治性前列腺切除术（RP）的男性。在 T2w MRI 上自动勾画前列腺 ROI，然后对其进行基于强度的采集伪影校正。一名专家病理学家在离体 RP 标本的染色切片上手动勾画主导肿瘤区域，并确定每个研究是 CG 还是 PZ CaP。使用非线性配准方案在空间上对齐，并将离体 RP 切片上的 CaP 范围映射到相应的 MRI 切片上。然后，从所有 T2w MRI 数据集上的每个体素上提取总共 110 个纹理特征。然后应用信息论特征选择过程来识别包含 CG 和 PZ CaP 各自特定 T2w MRI 纹理特征的 QIS。通过将从相应组织学映射而来的 T2w MRI 上的 CaP 作为ground truth，对 CG 和 PZ CaP 的 QIS 进行了基于体素的二次判别分析（QDA）评估。
在随机三折交叉验证的 25 次运行中，QDA 分类器在 CG CaP 研究中的受试者工作特征曲线下面积为 0.86，在 PZ CaP 研究中的受试者工作特征曲线下面积为 0.73。相比之下，当（a）使用所有 110 个纹理特征（未应用特征选择）以及（b）随机选择的纹理特征组合时，QDA 分类器的准确性明显降低。
在体内直肠内 T2w MRI 上，CG 和 PZ 前列腺癌具有明显不同的纹理定量成像特征。
