Awate Suyash P, Leahy Richard M, Joshi Anand A
Computer Science and Engineering Department, Indian Institute of Technology (IIT) Bombay, Mumbai, India.
Signal and Image Processing Institute (SIPI), University of Southern California, Los Angeles, USA.
Inf Process Med Imaging. 2017 Jun;10265:28-40. doi: 10.1007/978-3-319-59050-9_3. Epub 2017 May 23.
Typical cerebral cortical analyses rely on spatial normalization and are sensitive to misregistration arising from partial homologies between subject brains and local optima in nonlinear registration. In contrast, we use a descriptor of the 3D cortical sheet (jointly modeling folding and thickness) that is robust to misregistration. Our histogram-based descriptor lies on a . We propose new for (i) detecting group differences, using a with an implicit lifting map to a reproducing kernel Hilbert space, and (ii) regression against clinical variables, using . For both methods, we employ kernels that exploit the Riemannian structure. Results on simulated and clinical data shows the improved accuracy and stability of our approach in cortical-sheet analysis.
典型的大脑皮层分析依赖于空间归一化，并且对由于个体大脑之间的部分同源性以及非线性配准中的局部最优解而产生的配准错误很敏感。相比之下，我们使用一种三维皮质层的描述符（联合对折叠和厚度进行建模），它对配准错误具有鲁棒性。我们基于直方图的描述符位于一个……上。我们提出了新的方法用于：（i）检测组间差异，使用带有到再生核希尔伯特空间的隐式提升映射的……，以及（ii）针对临床变量进行回归，使用……。对于这两种方法，我们都采用利用黎曼结构的核。模拟数据和临床数据的结果表明了我们的方法在皮质层分析中具有更高的准确性和稳定性。
