Javier Villanueva-Meyer

Pulsatile tinnitus, a severely debilitating condition where a “whooshing” sound is heard with each heart beat, affects over 3 million Americans, leading to anxiety, depression and even suicide. While many there are many diseases associated with pulsatile tinnitus one common and serious condition is idiopathic intracranial hypertension. Our study is the first to show that the presence of brain herniations through arachnoid granulations into dural sinuses on MRI is significantly associated with idiopathic intracranial hypertension in patient's with pulsatile tinnitus.

This Editorial Comment discusses the following AJR article: Association of Time Since Administration of Pegylated G-CSF (Pegfilgrastim) and Bone Marrow Uptake on FDG PET/CT: Determination of a Mini...

Survival outcomes for patients with lower grade gliomas (LrGG) are improving. However, injury from tumor growth and consequences of treatment often leads to impaired cognition, particularly in cognitive domains reliant on distributed functional networks and intact white-matter tracts. Resting state functional MRI (rsfMRI) is a method of investigating the integrity of these functional networks. This study investigated rsfMRI connectivity in 21 patients with clinically stable LrGG compared to age- and gender-matched healthy controls, and associated imaging measures with cognitive outcomes. Data were acquired for 12 cognitive tests administered within one week of imaging. RsfMRI and T1-weighted images for 21 research controls were acquired from OpenNeuro datasets. RsfMRI data were processed and analyzed using the CONN toolbox using CONN’s standard regions of interest (ROI) for the 8 canonical networks as seeds, and cognitive test scores as covariates, with a threshold for T tests of p< .001 uncorrected. Median age was 48 years old (range 27-67). There were 6 astrocytomas, IDHmut; 3 astrocytomas IDH-wt, 8 oligodendrogliomas, and 4 NOS. Thirteen had left hemisphere tumors (8 frontal, 3 parietal, 2 temporal), and 6 right (5 frontal, 1 temporal). Fourteen had previously recieved radiotherapy. There was significantly lower connectivity in frontoparietal executive control and the salience networks in LrGG patients versus controls. Within patients, lower executive control network connectivity covaried with worse performance on executive measures (FAS, Tower of London, Trails-A, Animal Naming, FrSBe), and attention and working memory measures (Digit Symbol, HVLT). Lower salience network connectivity covaried with poorer performance on executive measures (FrSBe, FAS) and attention and working memory measures (Digit Span, HVLT, WAIS-WM). In clinically stable LrGG, rsfMRI measures of network connectivity are potentially useful markers to monitor and track, given the concordance with cognition, and could help guide cognitive assessment and rehabilitation.

Several nuclear imaging techniques can be used to diagnose infectious and inflammatory processes. F-18-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) is a useful diagnostic technique to detect inflammation and infection quickly and accurately. We report the case of a patient with end-stage renal disease (ESRD) and recurrent bacterial infections where FDG PET/CT was used to identify the source of infection as sternal osteomyelitis and rule out suspected infection of the aortic valve prosthesis.

Volume calculations have not been adopted into glioma response assessment due to lengthy times for manual definition and unreliable measures provided by automated algorithms. Relatively new artificial intelligence approaches such as convolutional neural networks have significantly improved lesion segmentation with performance accuracies >90%. However, their adoption into routine practice remains limited due to poor generalizability and failure rates approaching 25% when incorporated into clinical workflow. The latter can be attributed to 1) the requirement of four different types of anatomic images (T2, T2-FLAIR, T1 pre- and post-contrast); 2) cumbersome preprocessing including alignment, reformatting, and skull removal; and 3) the lack of a well-integrated clinical deployment system. The goal of this study was to demonstrate how simple modifications to a robust network coupled with an integrated workflow can provide reliable measures of tumor volume for real-time use in the reading room. Leveraging NVIDIA’s Clara-Train software and a molecularly diverse dataset of 400 labeled images for training, we modified a top-performing ensembled 2D-U-Net to require a single image-volume input (T2-FLAIR or post-contrast T1 for the T2-hyperintense or contrast-enhancing lesions) and deployed the results in the clinic to provide quantitative volumetrics. Inference was performed on a mix of image orientations without any reformatting or skull-stripping. Training on only 115 of our 400 datasets, we achieved Dice Coefficients of 90% and 81% overlap of our auto-segmented T2 and contrast-enhancing lesions with manual labels in our 25-patient validation cohort (11 enhancing), compared to 91% and 83% overlap with the original model that required four anatomic images to segment each lesion. Radiologists can view segmentations directly from PACS as contours or overlays and provide numerical feedback for model refinement. The workflow has been applied on 50 cases to date without any failures and can be easily shared for deployment on any clinical PACS.