Buchsbaum M S, Fallon J H, Wei T C, Guich S, Spiegel-Cohen J, Hamilton M, Tang C
Department of Psychiatry, Mount Sinai School of Medicine, New York, NY 10029-6574, USA.
Psychiatry Res. 1998 Dec 14;84(2-3):113-25. doi: 10.1016/s0925-4927(98)00052-3.
Functional as well as structural assessment of the basal forebrain has mostly focused on the dorsal caudate and putamen in axial slices where they are easily outlined or their centers located with stereotaxic methods. The more ventral extent of the basal forebrain, where the irregular form and indistinct boundaries of the nucleus accumbens and substantia innominata are difficult to trace and where the brain's ventral surface may contribute partial volume artifacts to measurement, has been less studied. We present a method based on coronal sections, landmarks placed on clearly visible anchor points, and the computational technique of thin-plate spline warping which allows the alignment of groups of individuals to common coordinates for pixel-by-pixel statistical mapping. The reliability of the landmarks across independent raters yields a median absolute difference of 1.3-1.6 mm. The validity of the method is confirmed by variance maps which reveal significant decreases in variance over spindle and bounding box alignment.
对基底前脑的功能和结构评估大多集中在轴位切片中的背侧尾状核和壳核，在这些切片中它们很容易勾勒出来，或者可以通过立体定向方法确定其中心位置。基底前脑更靠腹侧的部分，其中伏隔核和无名质的不规则形状和不清晰边界难以追踪，并且大脑腹侧表面可能会对测量产生部分容积伪影，这方面的研究较少。我们提出了一种基于冠状切片、放置在清晰可见锚点上的地标以及薄板样条扭曲计算技术的方法，该技术允许将个体组对齐到共同坐标以进行逐像素统计映射。独立评分者之间地标可靠性产生的中位数绝对差为1.3 - 1.6毫米。方差图证实了该方法的有效性，方差图显示在纺锤体和边界框对齐后方差显著降低。
