In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences. Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients—manually annotated by up to four raters—and to 65 comparable scans generated using tumor image simulation software. Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%–85%), illustrating the difficulty of this task. We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously. Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements. The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource.

Gliomas belong to a group of central nervous system tumors, and consist of various sub-regions. Gold standard labeling of these sub-regions in radiographic imaging is essential for both clinical and computational studies, including radiomic and radiogenomic analyses. Towards this end, we release segmentation labels and radiomic features for all pre-operative multimodal magnetic resonance imaging (MRI) (n=243) of the multi-institutional glioma collections of The Cancer Genome Atlas (TCGA), publicly available in The Cancer Imaging Archive (TCIA). Pre-operative scans were identified in both glioblastoma (TCGA-GBM, n=135) and low-grade-glioma (TCGA-LGG, n=108) collections via radiological assessment. The glioma sub-region labels were produced by an automated state-of-the-art method and manually revised by an expert board-certified neuroradiologist. An extensive panel of radiomic features was extracted based on the manually-revised labels. This set of labels and features should enable i) direct utilization of the TCGA/TCIA glioma collections towards repeatable, reproducible and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments, as well as ii) performance evaluation of computer-aided segmentation methods, and comparison to our state-of-the-art method.

Object. Because lumbar magnetic resonance (MR) imaging fails to identify a treatable cause of chronic sciatica in nearly 1 million patients annually, the authors conducted MR neurography and interventional MR imaging in 239 consecutive patients with sciatica in whom standard diagnosis and treatment failed to effect improvement. Methods. After performing MR neurography and interventional MR imaging, the final rediagnoses included the following: piriformis syndrome (67.8%), distal foraminal nerve root entrapment (6%), ischial tunnel syndrome (4.7%), discogenic pain with referred leg pain (3.4%), pudendal nerve entrapment with referred pain (3%), distal sciatic entrapment (2.1%), sciatic tumor (1.7%), lumbosacral plexus entrapment (1.3%), unappreciated lateral disc herniation (1.3%), nerve root injury due to spinal surgery (1.3%), inadequate spinal nerve root decompression (0.8%), lumbar stenosis (0.8%), sacroiliac joint inflammation (0.8%), lumbosacral plexus tumor (0.4%), sacral fracture (0.4%), and no diagnosis (4.2%). Open MR—guided Marcaine injection into the piriformis muscle produced the following results: no response (15.7%), relief of greater than 8 months (14.9%), relief lasting 2 to 4 months with continuing relief after second injection (7.5%), relief for 2 to 4 months with subsequent recurrence (36.6%), and relief for 1 to 14 days with full recurrence (25.4%). Piriformis surgery (62 operations; 3-cm incision, transgluteal approach, 55% outpatient; 40% with local or epidural anesthesia) resulted in excellent outcome in 58.5%, good outcome in 22.6%, limited benefit in 13.2%, no benefit in 3.8%, and worsened symptoms in 1.9%. Conclusions. This Class A quality evaluation of MR neurography's diagnostic efficacy revealed that piriformis muscle asymmetry and sciatic nerve hyperintensity at the sciatic notch exhibited a 93% specificity and 64% sensitivity in distinguishing patients with piriformis syndrome from those without who had similar symptoms (p < 0.01). Evaluation of the nerve beyond the proximal foramen provided eight additional diagnostic categories affecting 96% of these patients. More than 80% of the population good or excellent functional outcome was achieved.

The results of the challenge showed that the lungs and heart can be segmented fairly accurately by various algorithms, while deep-learning methods performed better on the esophagus. Our dataset together with the manual contours for all training cases continues to be available publicly as an ongoing benchmarking resource.

We have developed a prototype PET detector which is compatible with a clinical MRI system to provide simultaneous PET and MR imaging. This single-slice PET system consists of 48 2 x 2 x 10 mm3 LSO crystals in a 38 mm diameter ring configuration that can be placed inside the receiver coil of the MRI system, coupled to three multi-channel photomultipliers housed outside the main magnetic field via 4 m long and 2 mm diameter optical fibres. The PET system exhibits 2 mm spatial resolution, 41% energy resolution at 511 keV and 20 ns timing resolution. Simultaneous PET and MR phantom images were successfully acquired.

† People involved in the organization of the challenge. ‡ People contributing data from their institutions.§ Equal senior authors.

Targeted, noninvasive neuromodulation of the brain of an otherwise awake subject could revolutionize both basic and clinical neuroscience. Toward this goal, we have developed nanoparticles that allow noninvasive uncaging of a neuromodulatory drug, in this case the small molecule anesthetic propofol, upon the application of focused ultrasound. These nanoparticles are composed of biodegradable and biocompatible constituents and are activated using sonication parameters that are readily achievable by current clinical transcranial focused ultrasound systems. These particles are potent enough that their activation can silence seizures in an acute rat seizure model. Notably, there is no evidence of brain parenchymal damage or blood-brain barrier opening with their use. Further development of these particles promises noninvasive, focal, and image-guided clinical neuromodulation along a variety of pharmacological axes.

High-resolution microscopy images of tissue specimens provide detailed information about the morphology of normal and diseased tissue. Image analysis of tissue morphology can help cancer researchers develop a better understanding of cancer biology. Segmentation of nuclei and classification of tissue images are two common tasks in tissue image analysis. Development of accurate and efficient algorithms for these tasks is a challenging problem because of the complexity of tissue morphology and tumor heterogeneity. In this paper we present two computer algorithms; one designed for segmentation of nuclei and the other for classification of whole slide tissue images. The segmentation algorithm implements a multiscale deep residual aggregation network to accurately segment nuclear material and then separate clumped nuclei into individual nuclei. The classification algorithm initially carries out patch-level classification via a deep learning method, then patch-level statistical and morphological features are used as input to a random forest regression model for whole slide image classification. The segmentation and classification algorithms were evaluated in the MICCAI 2017 Digital Pathology challenge. The segmentation algorithm achieved an accuracy score of 0.78. The classification algorithm achieved an accuracy score of 0.81. These scores were the highest in the challenge.