Previously titled: Advanced Multimodality Image Guided Operating Suite (AMIGO)

The Brain Resection Multimodal Imaging Database (ReMind) 

The standard of care for brain tumors is maximal safe surgical resection as the first step. Neuronavigation augments the surgeon’s ability to achieve this, but loses validity due to brain shift as surgery progresses. Moreover, many gliomas are difficult to distinguish from adjacent brain tissue.

Intraoperative MRI is a useful intraoperative adjunct which can be used to visualize residual tumor and brain shift. Intraoperative ultrasound is faster and easier to incorporate into the workflow, but provides lower contrast between tissue and normal brain tissue.

With the success of data-hungry AI/ML algorithms in advancing the state of the art in medical image analysis, the benefits of sharing well curated data can not be overstated.

To this end, we provide here the largest publicly-available MRI and intraoperative ultrasound imaging database of surgically treated brain tumors, including gliomas (n=92), metastases (n=11) and others (n=11). This collection contains 369 preoperative MRI series, 320 3D intraoperative ultrasound series, 300 intraoperative MRI series, and 358 segmentations collected from 114 consecutive patients at a single institution. We expect this data to be a resource for computational research in brain shift and image analysis as well as for training in interpretation of intraoperative ultrasound and MRI for neurosurgery.

To the best of our knowledge there are no other intraoperative brain tumor resection MRI and Ultrasound datasets in TCIA. Understanding brain shift, and accounting for it during brain tumor resection is an open problem, and this dataset may help computer vision researchers develop algorithms for image segmentation and registration to help understand how the brain shifts and deforms during surgery.