BACKGROUND AND PURPOSE:The World Health Organization has recently placed new emphasis on the integration of genetic information for gliomas. While tissue sampling remains the criterion standard, noninvasive imaging techniques may provide complimentary insight into clinically relevant genetic mutations. Our aim was to train a convolutional neural network to independently predict underlying molecular genetic mutation status in gliomas with high accuracy and identify the most predictive imaging features for each mutation.

A customized deep learning tool is accurate in the detection and quantification of hemorrhage on NCCT. Demonstrated high performance on prospective NCCTs ordered from the emergency department suggests the clinical viability of the proposed deep learning tool.

Background: Computer-aided methods have been widely applied to diagnose lesions detected on breast MRI, but fullyautomatic diagnosis using deep learning is rarely reported. Purpose: To evaluate the diagnostic accuracy of mass lesions using region of interest (ROI)-based, radiomics and deeplearning methods, by taking peritumor tissues into consideration. Study Type: Retrospective. Population: In all, 133 patients with histologically confirmed 91 malignant and 62 benign mass lesions for training (74 patients with 48 malignant and 26 benign lesions for testing). Field Strength/Sequence: 3T, using the volume imaging for breast assessment (VIBRANT) dynamic contrast-enhanced (DCE) sequence. Assessment: 3D tumor segmentation was done automatically by using fuzzy-C-means algorithm with connectedcomponent labeling. A total of 99 texture and histogram parameters were calculated for each case, and 15 were selected using random forest to build a radiomics model. Deep learning was implemented using ResNet50, evaluated with 10-fold crossvalidation. The tumor alone, smallest bounding box, and 1.2, 1.5, 2.0 times enlarged boxes were used as inputs. Statistical Tests: The malignancy probability was calculated using each model, and the threshold of 0.5 was used to make a diagnosis. Results: In the training dataset, the diagnostic accuracy was 76% using three ROI-based parameters, 84% using the radiomics model, and 86% using ROI + radiomics model. In deep learning using the per-slice basis, the area under the receiver operating characteristic (ROC) was comparable for tumor alone, smallest and 1.2 times box (AUC = 0.97-0.99), which were significantly higher than 1.5 and 2.0 times box (AUC = 0.86 and 0.71, respectively). For per-lesion diagnosis, the highest accuracy of 91% was achieved when using the smallest bounding box, and that decreased to 84% for tumor alone and 1.2 times box, and further to 73% for 1.5 times box and 69% for 2.0 times box. In the independent testing dataset, the perlesion diagnostic accuracy was also the highest when using the smallest bounding box, 89%. Data Conclusion: Deep learning using ResNet50 achieved a high diagnostic accuracy. Using the smallest bounding box containing proximal peritumor tissue as input had higher accuracy compared to using tumor alone or larger boxes. Level of Evidence: 3 Technical Efficacy: Stage 2

ABBREVIATIONS 5-ALA = 5-aminolevulinic acid; BBB = blood-brain barrier; CE = contrast enhancing; CUMC = Columbia University Medical Center; EOR = extent of resection; GBM = glioblastoma; GTR = gross-total resection; HGG = high-grade glioma; n-ANOVA = multiway ANOVA; NCE = non-contrast enhancing; OS = overall survival; PFS = progression-free survival; PPV = positive predictive value; ROI = region of interest; STR = subtotal resection; UD = undetermined; WHO = World Health Organization; Y560 = YELLOW 560. OBJECTIVE Extent of resection is an important prognostic factor in patients undergoing surgery for glioblastoma (GBM). Recent evidence suggests that intravenously administered fluorescein sodium associates with tumor tissue, facilitating safe maximal resection of GBM. In this study, the authors evaluate the safety and utility of intraoperative fluorescein guidance for the prediction of histopathological alteration both in the contrast-enhancing (CE) regions, where this relationship has been established, and into the non-CE (NCE), diffusely infiltrated margins. METHODS Thirty-two patients received fluorescein sodium (3 mg/kg) intravenously prior to resection. Fluorescence was intraoperatively visualized using a Zeiss Pentero surgical microscope equipped with a YELLOW 560 filter. Stereotactically localized biopsy specimens were acquired from CE and NCE regions based on preoperative MRI in conjunction with neuronavigation. The fluorescence intensity of these specimens was subjectively classified in real time with subsequent quantitative image analysis, histopathological evaluation of localized biopsy specimens, and radiological volumetric assessment of the extent of resection. RESULTS Bright fluorescence was observed in all GBMs and localized to the CE regions and portions of the NCE margins of the tumors, thus serving as a visual guide during resection. Gross-total resection (GTR) was achieved in 84% of the patients with an average resected volume of 95%, and this rate was higher among patients for whom GTR was the surgical goal (GTR achieved in 93.1% of patients, average resected volume of 99.7%). Intraoperative fluorescein staining correlated with histopathological alteration in both CE and NCE regions, with positive predictive values by subjective fluorescence evaluation greater than 96% in NCE regions. CONCLUSIONS Intraoperative administration of fluorescein provides an easily visualized marker for glioma pathology in both CE and NCE regions of GBM. These findings support the use of fluorescein as a microsurgical adjunct for guiding GBM resection to facilitate safe maximal removal. SUBMITTEDhttps://thejns.org/doi/abs/10.3171/2016.7.JNS16232

Artificial intelligence technology is a rapidly expanding field with many applications in acute stroke imaging, including ischemic and hemorrhage subtypes. Early identification of acute stroke is critical for initiating prompt intervention to reduce morbidity and mortality. Artificial intelligence can help with various aspects of the stroke treatment paradigm, including infarct or hemorrhage detection, segmentation, classification, large vessel occlusion detection, Alberta Stroke Program Early CT Score grading, and prognostication. In particular, emerging artificial intelligence techniques such as convolutional neural networks show promise in performing these imaging-based tasks efficiently and accurately. The purpose of this review is twofold: first, to describe AI methods and available public and commercial platforms in stroke imaging, and second, to summarize the literature of current artificial intelligence-driven applications for acute stroke triage, surveillance, and prediction. ABBREVIATIONS: AI ¼ artificial intelligence; ANN ¼ artificial neural network; AUC ¼ area under the curve; CNN ¼ convolutional neural network; DL ¼ deep learning; ICC ¼ intraclass correlation coefficient; ICH ¼ intracranial hemorrhage; LVO ¼ large vessel occlusion; ML ¼ machine learning; MRP ¼ MR perfusion; RF ¼ random forest; SVM ¼ support vector machine S troke is the second leading cause of death worldwide with an annual mortality of about 5.5 million. 1,2 In the United States, nearly 800,000 people have a stroke annually, and the economic burden of stroke is estimated at $34 billion per year. 3 Morbidity is high, with more than half of patients with stroke left chronically disabled. 2 Neuroimaging is an important tool for the detection, characterization, and prognostication of acute strokes, including ischemic and hemorrhagic subtypes. Artificial intelligence (AI) technology is a rapidly burgeoning field, providing a promising avenue for fast and efficient imaging analysis. 4 AI applications for imaging of acute cerebrovascular disease have been implemented, including tools for triage, quantification, surveillance, and prediction. This review aims to summarize the current landscape of AIdriven applications for acute cerebrovascular disease assessment focusing primarily on deep learning (DL) methods. OVERVIEW OF AI Although AI, machine learning (ML), and DL are used interchangeably, these in fact represent subdisciplines. Specifically, DL is a subset of ML, and ML is a subset of AI (Fig 1). Broadly, AI uses computers to perform tasks that typically require human knowledge. ML, a subset of AI, uses statistical approaches to enable machines to optimize outcome prediction as they are exposed to data and train computers for pattern recognition, a task generally requiring human intelligence. 5 ML offers several potential advantages over visual inspection by human experts, including objective and quantitative evaluation, the ability to detect subtle voxel-level patterns, speed, and large-scale implementation. Feature selection, classifi...