A method for the automatic segmentation of brown adipose tissue.

OBJECTIVE: Brown adipose tissue (BAT) plays a key role for thermogenesis in mammals and infants. Recent confirmation of BAT presence in adult humans has aroused great interest for its potential to initiate weight-loss and normalize metabolic disorders in diabetes and obesity. Reliable detection and differentiation of BAT from the surrounding white adipose tissue (WAT) and muscle is critical for assessment/quantification of BAT volume. This study evaluates magnetic resonance (MR) acquisition for BAT and the efficacy of different automated methods for MR features-based BAT segmentation to identify the best suitable method. MATERIALS AND METHODS: Multi-point Dixon and multi-echo T2 spin-echo images were acquired from 12 mice using an Agilent 9.4T scanner. Four segmentation methods: multidimensional thresholding (MTh); region-growing (RG); fuzzy c-means (FCM) and neural-network (NNet) were evaluated for the interscapular region and validated against manually defined BAT, WAT and muscle. RESULTS: Statistical analysis of BAT segmentation yielded a median Dice-Statistical-Index, and sensitivity of 89.92% for NNet, 82.86% for FCM, 72.74% for RG, and 72.70%, for MTh, respectively. CONCLUSION: This study demonstrates that NNet improves the specificity to BAT from surrounding tissue based on 3-point Dixon and T2 MRI. This method facilitates quantification and longitudinal measurement of BAT in preclinical-models and human subjects.

An end-end deep learning framework for lesion segmentation on multi-contrast MR images-an exploratory study in a rat model of traumatic brain injury.

Traumatic brain injury (TBI) engenders traumatic necrosis and penumbra-areas of secondary neural injury which are crucial targets for therapeutic interventions. Segmenting manually areas of ongoing changes like necrosis, edema, hematoma, and inflammation is tedious, error-prone, and biased. Using the multi-parametric MR data from a rodent model study, we demonstrate the effectiveness of an end-end deep learning global-attention-based UNet (GA-UNet) framework for automatic segmentation and quantification of TBI lesions. Longitudinal MR scans (2 h, 1, 3, 7, 14, 30, and 60 days) were performed on eight Sprague-Dawley rats after controlled cortical injury was performed. TBI lesion and sub-regions segmentation was performed using 3D-UNet and GA-UNet. Dice statistics (DSI) and Hausdorff distance were calculated to assess the performance. MR scan variations-based (bias, noise, blur, ghosting) data augmentation was performed to develop a robust model.Training/validation median DSI for U-Net was 0.9368 with T2w and MPRAGE inputs, whereas GA-UNet had 0.9537 for the same. Testing accuracies were higher for GA-UNet than U-Net with a DSI of 0.8232 for the T2w-MPRAGE inputs.Longitudinally, necrosis remained constant while oligemia and penumbra decreased, and edema appearing around day 3 which increased with time. GA-UNet shows promise for multi-contrast MR image-based segmentation/quantification of TBI in large cohort studies.

Amyloid pathology fingerprint differentiates post-traumatic stress disorder and traumatic brain injury.

Introduction: Traumatic brain injury (TBI) and post-traumatic stress disorder (PTSD) are risk factors for early onset of Alzheimer's disease (AD) and may accelerate the progression rate of AD pathology. As amyloid-beta (Aß) plaques are a hallmark of AD pathology, we hypothesized that TBI and PTSD might increase Aß accumulation in the brain. Methods: We examined PET and neuropsychological data from Vietnam War veterans compiled by the US Department of Defense Alzheimer's Disease Neuroimaging Initiative, to examine the spatial distribution of Aß in male veterans' who had experienced a TBI and/or developed PTSD. Subjects were classified into controls, TBI only, PTSD only, and TBI with PTSD (TBI_PTSD) groups and data were analyzed using both voxel-based and ROI-based approaches. Results: Compared to controls, all three clinical groups showed a pattern of mainly increased referenced standard uptake values (SUVR) for the amyloid tracer [18F]-AV45 PET, with rank order PTSD > TBI_PTSD > TBI > Control, and same rank order was seen in the deficits of cognitive functions. SUVR increase was observed in widespread cortical regions of the PTSD group; in white matter of the TBI_PTSD group; and cerebellum and precuneus area of the TBI group, in contrast with controls. The [18F]-AV45 SUVR correlated negatively with cerebrospinal fluid (CSF) amyloid levels and positively with the CSF tau concentrations. Conclusion: These results suggest that both TBI and PTSD are substantial risk factors for cognition decline and increased Aß deposition resembling that in AD. In addition, both PTSD and TBI_PTSD have a different pathways of Aß accumulation.