McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal H3A 2B4, Canada.
Med Phys. 2012 Jun;39(6):3253-61. doi: 10.1118/1.4709600.
One of the important challenges in the field of medical imaging is finding real clinical images with which to validate new image processing algorithms. This is particularly true for tracked 3D ultrasound images of the brain.
In 2010, pre- and postoperative magnetic resonance and intraoperative ultrasound images were acquired from brain tumor patients involved in the authors' imaging study at the Montreal Neurological Institute.
These data are available online at the Montreal Neurological Institute's Brain Images of Tumors for Evaluation database, termed here the MNI BITE database. It contains ultrasound and magnetic resonance images from 14 patients. Each patient underwent a preoperative and a postoperative T1-weighted magnetic resonance scan with gadolinium enhancement, and multiple intraoperative B-mode images were acquired before and after resection. Corresponding features were manually selected in some image pairs for validation. All images are in MINC format, the file format used at the authors' institute for image processing. The MINC tools are available for free download at packages.bic.mni.mcgill.ca.
This is the first online database of its kind. These images can be used by image processing scientists as well as clinicians wishing to compare findings from magnetic resonance and ultrasound imaging.
医学影像学领域的一个重要挑战是寻找真实的临床图像来验证新的图像处理算法。对于跟踪式 3D 脑超声图像来说尤其如此。
2010 年，作者在蒙特利尔神经学研究所的成像研究中收集了脑肿瘤患者的术前和术后磁共振成像和术中超声图像。
这些数据可在蒙特利尔神经学研究所的脑肿瘤评估图像数据库（MNI BITE 数据库）中在线获取。该数据库包含 14 名患者的超声和磁共振图像。每位患者都接受了术前和术后钆增强 T1 加权磁共振扫描，并在切除前和切除后采集了多次术中 B 模式图像。在一些图像对中手动选择了相应的特征用于验证。所有图像均采用 MINC 格式，这是作者所在机构用于图像处理的文件格式。MINC 工具可在 packages.bic.mni.mcgill.ca 免费下载。
这是第一个此类在线数据库。图像处理科学家和希望比较磁共振成像和超声成像结果的临床医生都可以使用这些图像。
