Machine learning is a powerful technique for recognizing patterns on medical images; however, it must be used with caution because it can be misused if the strengths and weaknesses of this technolo... Machine learning is a technique for recognizing patterns that can be applied to medical images. Although it is a powerful tool that can help in rendering medical diagnoses, it can be misapplied. Ma...

Quantitative analysis of brain MRI is routine for many neurological diseases and conditions and relies on accurate segmentation of structures of interest. Deep learning-based segmentation approaches for brain MRI are gaining interest due to their self-learning and generalization ability over large amounts of data. As the deep learning architectures are becoming more mature, they gradually outperform previous state-of-the-art classical machine learning algorithms. This review aims to provide an overview of current deep learning-based segmentation approaches for quantitative brain MRI. First we review the current deep learning architectures used for segmentation of anatomical brain structures and brain lesions. Next, the performance, speed, and properties of deep learning approaches are summarized and discussed. Finally, we provide a critical assessment of the current state and identify likely future developments and trends.

This summary of the 2018 NIH/RSNA/ACR/The Academy Workshop on Artificial Intelligence in Medical Imaging provides a roadmap to identify and prioritize research needs for academic research laborator... Imaging research laboratories are rapidly creating machine learning systems that achieve expert human performance using open-source methods and tools. These artificial intelligence systems are bein...

Fully automated and two-dimensional abdominal segmentation of CT scans performed using a deep-convolutional neural network for assessment of body composition. PurposeTo develop and evaluate a fully automated algorithm for segmenting the abdomen from CT to quantify body composition.Materials and MethodsFor this retrospective study, a convolutional neural network based on the U-Net architecture was trained to perform abdominal segmentation on a data set of 2430 two-dimensional CT examinations and was tested on 270 CT examinations. It was further tested on a separate data set of 2369 patients with hepatocellular carcinoma (HCC). CT examinations were performed between 1997 and 2015. The mean age of patients was 67 years; for male patients, it was 67 years (range, 29–94 years), and for female patients, it was 66 years (range, 31–97 years). Differences in segmentation performance were assessed by using two-way analysis of variance with Bonferroni correction.ResultsCompared with reference segmentation, the model for this study achieved Dice scores (mean ± standard deviation) of 0.98 ± 0.03, 0.96 ± 0.02, and 0.97 ± 0.01 in the test set, and 0.94 ± 0.05, 0.92 ± 0.04, an...

Purpose: The purpose of this article is to determine the response rate and toxicity of PCV administered before radiation therapy in patients with newly diagnosed LGO/LGOA and to explore correlations between response with 1p/19q deletions and aberrant p53 expression. Background: Despite prolonged survival of patients with low-grade oligodendroglioma (LGO) and oligoastrocytoma (LGOA), the majority will succumb to progressive disease. Because procarbazine, lomustine (CCNU), and vincristine (PCV) is active in patients with recurrent LGO/LGOA, we hypothesized that it would be beneficial as primary therapy. Methods: Adult patients with residual tumor on magnetic resonance imaging scan following biopsy or subtotal resection of LGO/LGOA received up to six cycles of PCV. Radiation therapy (59.4 or 54.0 Gy) began within 10 weeks of completing chemotherapy or immediately if there was evidence of tumor progression on PCV. Tumor tissue was analyzed by fluorescent in situ hybridization for 1p and 19q deletion and by immunohistochemistry for p53 expression. Results: Eight of 28 (29%) and 13 of 25 (52%) eligible patients demonstrated tumor regression as assessed by the treating physician and a blinded central neuroradiology reviewer, respectively. Myelosuppression was the predominant toxicity. Loss of 1p and 19q were associated with LGO but not LGOA (P = .009), were inversely associated with p53 detection, and were not associated with response to PCV (possibly because of the small sample size). Conclusion: PCV produces tumor regressions in a meaningful proportion of patients with LGO/LGOA. Toxicity, especially myelosuppression, is significant. Loss of 1p and 19q seems limited to patients with pure LGO and is inversely related to p53 alterations.