Interrelationships between Cellular Density,Mosaic Patterning,and Dendritic Coverage of VGluT3 Amacrine Cells

Amacrine cells of the retina are conspicuously variable in their morphologies, their population demographics, and their ensuing functions. Vesicular glutamate transporter 3 (VGluT3) amacrine cells are a recently characterized type of amacrine cell exhibiting local dendritic autonomy. The present analysis has examined three features of this VGluT3 population, including their density, local distribution, and dendritic spread, to discern the extent to which these are interrelated, using male and female mice. We first demonstrate that Bax-mediated cell death transforms the mosaic of VGluT3 cells from a random distribution into a regular mosaic. We subsequently examine the relationship between cell density and mosaic regularity across recombinant inbred strains of mice, finding that, although both traits vary across the strains, they exhibit minimal covariation. Other genetic determinants must therefore contribute independently to final cell number and to mosaic order. Using a conditional KO approach, we further demonstrate that Bax acts via the bipolar cell population, rather than cell-intrinsically, to control VGluT3 cell number. Finally, we consider the relationship between the dendritic arbors of single VGluT3 cells and the distribution of their homotypic neighbors. Dendritic field area was found to be independent of Voronoi domain area, while dendritic coverage of single cells was not conserved, simply increasing with the size of the dendritic field. Bax-KO retinas exhibited a threefold increase in dendritic coverage. Each cell, however, contributed less dendrites at each depth within the plexus, intermingling their processes with those of neighboring cells to approximate a constant volumetric density, yielding a uniformity in process coverage across the population.SIGNIFICANCE STATEMENT Different types of retinal neuron spread their processes across the surface of the retina to achieve a degree of dendritic coverage that is characteristic of each type. Many of these types achieve a constant coverage by varying their dendritic field area inversely with the local density of like-type neighbors. Here we report a population of retinal amacrine cells that do not develop dendritic arbors in relation to the spatial positioning of such homotypic neighbors; rather, this cell type modulates the extent of its dendritic branching when faced with a variable number of overlapping dendritic fields to approximate a uniformity in dendritic density across the retina.     

A Stacked Generalization of 3D Orthogonal Deep Learning Convolutional Neural Networks for Improved Detection of White Matter Hyperintensities in 3D FLAIR Images

BACKGROUND AND PURPOSE:Accurate and reliable detection of white matter hyperintensities and their volume quantification can provide valuable clinical information to assess neurologic disease progression. In this work, a stacked generalization ensemble of orthogonal 3D convolutional neural networks, StackGen-Net, is explored for improving automated detection of white matter hyperintensities in 3D T2-FLAIR images.MATERIALS AND METHODS:Individual convolutional neural networks in StackGen-Net were trained on 2.5D patches from orthogonal reformatting of 3D-FLAIR (n = 21) to yield white matter hyperintensity posteriors. A meta convolutional neural network was trained to learn the functional mapping from orthogonal white matter hyperintensity posteriors to the final white matter hyperintensity prediction. The impact of training data and architecture choices on white matter hyperintensity segmentation performance was systematically evaluated on a test cohort (n = 9). The segmentation performance of StackGen-Net was compared with state-of-the-art convolutional neural network techniques on an independent test cohort from the Alzheimer’s Disease Neuroimaging Initiative-3 (n = 20).RESULTS:StackGen-Net outperformed individual convolutional neural networks in the ensemble and their combination using averaging or majority voting. In a comparison with state-of-the-art white matter hyperintensity segmentation techniques, StackGen-Net achieved a significantly higher Dice score (0.76 [SD, 0.08], F1-lesion (0.74 [SD, 0.13]), and area under precision-recall curve (0.84 [SD, 0.09]), and the lowest absolute volume difference (13.3% [SD, 9.1%]). StackGen-Net performance in Dice scores (median = 0.74) did not significantly differ (P = .22) from interobserver (median = 0.73) variability between 2 experienced neuroradiologists. We found no significant difference (P = .15) in white matter hyperintensity lesion volumes from StackGen-Net predictions and ground truth annotations.CONCLUSIONS:A stacked generalization of convolutional neural networks, utilizing multiplanar lesion information using 2.5D spatial context, greatly improved the segmentation performance of StackGen-Net compared with traditional ensemble techniques and some state-of-the-art deep learning models for 3D-FLAIR.

White matter hyperintensities (WMHs) correspond to pathologic features of axonal degeneration, demyelination, and gliosis observed within cerebral white matter.¹ Clinically, the extent of WMHs in the brain has been associated with cognitive impairment, Alzheimer’s disease and vascular dementia, and increased risk of stroke.^2,3 The detection and quantification of WMH volumes to monitor lesion burden evolution and its correlation with clinical outcomes have been of interest in clinical research.^4,5 Although the extent of WMHs can be visually scored,⁶ the categoric nature of such scoring systems makes quantitative evaluation of disease progression difficult. Manually segmenting WMHs is tedious, prone to inter- and intraobserver variability, and is, in most cases, impractical. Thus, there is an increased interest in developing fast, accurate, and reliable computer-aided automated techniques for WMH segmentation.Convolutional neural network (CNN)-based approaches have been successful in several semantic segmentation tasks in medical imaging.⁷ Recent works have proposed using deep learning–based methods for segmenting WMHs using 2D-FLAIR images.^8-11 More recently, a WMH segmentation challenge¹² was also organized (http://wmh.isi.uu.nl/) to facilitate comparison of automated segmentation of WMHs of presumed vascular origin in 2D multislice T2-FLAIR images. Architectures that used an ensemble of separately trained CNNs showed promising results in this challenge, with 3 of the top 5 winners using ensemble-based techniques.¹²Conventional 2D-FLAIR images are typically acquired with thick slices (3–4 mm) and possible slice gaps. Partial volume effects from a thick slice are likely to affect the detection of smaller lesions, both in-plane and out-of-plane. 3D-FLAIR images, with isotropic resolution, have been shown to achieve higher resolution and contrast-to-noise ratio¹³ and have shown promising results in MS lesion detection using 3D CNNs.¹⁴ Additionally, the isotropic resolution enables viewing and evaluation of the images in multiple planes. This multiplanar reformatting of 3D-FLAIR without the use of interpolating kernels is only possible due to the isotropic nature of the acquisition. Network architectures that use information from the 3 orthogonal views have been explored in recent works for CNN-based segmentation of 3D MR imaging data.¹⁵ The use of data from multiple planes allows more spatial context during training without the computational burden associated with full 3D training.¹⁶ The use of 3 orthogonal views simultaneously mirrors how humans approach this segmentation task.Ensembles of CNNs have been shown to average away the variances in the solution and the choice of model- and configuration-specific behaviors of CNNs.¹⁷ Traditionally, the solutions from these separately trained CNNs are combined by averaging or using a majority consensus. In this work, we propose the use of a stacked generalization framework (StackGen-Net) for combining multiplanar lesion information from 3D CNN ensembles to improve the detection of WMH lesions in 3D-FLAIR. A stacked generalization¹⁸ framework learns to combine solutions from individual CNNs in the ensemble. We systematically evaluated the performance of this framework and compared it with traditional ensemble techniques, such as averaging or majority voting, and state-of-the-art deep learning techniques.     

Improved Diagnostic Sensitivity of Bowel Disease of Prematurity on Contrast-Enhanced Ultrasound

Bowel diseases of prematurity, including necrotizing enterocolitis, are dreaded ailments of neonates. Early diagnosis is difficult, with clinical and radiographic findings often inconclusive. We present a novel use of contrast-enhanced ultrasound in detection of pediatric bowel disease. Early identification of compromised blood flow or an at-risk bowel can be quantitatively detected and monitored. This ability has implications for guidance of emerging therapies, allowing targeting of inflammation. These findings represent an advancement in detection of bowel disease in neonates.     

Checkpoint-blocker-induced autoimmunity is associated with favourable outcome in metastatic melanoma and distinct T-cell expression profiles

Background Immune checkpoint blockers (ICBs) activate CD8⁺ T cells, eliciting both anti-cancer activity and immune-related adverse events (irAEs). The relationship of irAEs with baseline parameters and clinical outcome is unclear.Methods Retrospective evaluation of irAEs on survival was performed across primary (N = 144) and secondary (N = 211) independent cohorts of patients with metastatic melanoma receiving single agent (pembrolizumab/nivolumab—sICB) or combination (nivolumab and ipilimumab—cICB) checkpoint blockade. RNA from pre-treatment and post-treatment CD8⁺ T cells was sequenced and differential gene expression according to irAE development assessed.Results 58.3% of patients developed early irAEs and this was associated with longer progression-free (PFS) and overall survival (OS) across both cohorts (log-rank test, OS: P < 0.0001). Median survival for patients without irAEs was 16.6 months (95% CI: 10.9–33.4) versus not-reached (P = 2.8 × 10⁻⁶). Pre-treatment monocyte and neutrophil counts, but not BMI, were additional predictors of clinical outcome. Differential expression of numerous gene pathway members was observed in CD8⁺ T cells according to irAE development, and patients not developing irAEs demonstrating upregulated CXCR1 pre- and post-treatment.Conclusions Early irAE development post-ICB is associated with favourable survival in MM. Development of irAEs is coupled to expression of numerous gene pathways, suggesting irAE development in-part reflects baseline immune activation.Subject terms: Immunotherapy, Melanoma     

Characterization of Benign Myocarditis Using Quantitative Delayed-Enhancement Imaging Based on Molli T1 Mapping

Delayed contrast enhancement after injection of a gadolinium-chelate (Gd-chelate) is a reference imaging method to detect myocardial tissue changes. Its localization within the thickness of the myocardial wall allows differentiating various pathological processes such as myocardial infarction (MI), inflammatory myocarditis, and cardiomyopathies. The aim of the study was first to characterize benign myocarditis using quantitative delayed-enhancement imaging and then to investigate whether the measure of the extracellular volume fraction (ECV) can be used to discriminate between MI and myocarditis.In 6 patients with acute benign myocarditis (32.2 ± 13.8 year-old, subepicardial late gadolinium enhancement [LGE]) and 18 patients with MI (52.3 ± 10.9 year-old, subendocardial/transmural LGE), myocardial T1 was determined using the Modified Look-Locker Imaging (MOLLI) sequence at 3 Tesla before and after Gd-chelate injection. T1 values were compared in LGE and normal regions of the myocardium. The myocardial T1 values were normalized to the T1 of blood, and the ECV was calculated from T1 values of myocardium and blood pre- and post-Gd injection.In both myocarditis and MI, the T1 was lower in LGE regions than in normal regions of the left ventricle. T1 of LGE areas was significantly higher in myocarditis than in MI (446.8 ± 45.8 vs 360.5 ± 66.9 ms, P = 0.003) and ECV was lower in myocarditis than in MI (34.5 ± 3.3 vs 53.8 ± 13.0 %, P = 0.004).Both inflammatory process and chronic fibrosis induce LGE (subepicardial in myocarditis and subendocardial in MI). The present study demonstrates that the determination of T1 and ECV is able to differentiate the 2 histological patterns.Further investigation will indicate whether the severity of ECV changes might help refine the predictive risk of LGE in myocarditis.     

Interhemispheric visual competition after multisensory reversal of hemianopia

Unilateral lesions of visual cortex have the secondary consequence of suppressing visual circuits in the midbrain superior colliculus (SC), collectively producing blindness in contralesional space (“hemianopia”). Recent studies have demonstrated that SC visual responses and contralesional vision can be reinstated by a non‐invasive multisensory training procedure in which spatiotemporally concordant visual‐auditory pairs are repeatedly presented within the blind hemifield. Despite this recovery of visual responsiveness, the loss of visual cortex was expected to result in permanent deficits in that hemifield, especially when visual events in both hemifields compete for attention and access to the brain's visuomotor circuitry. This was evaluated in the present study in a visual choice paradigm in which the two visual hemifields of recovered cats were simultaneously stimulated with equally valent visual targets. Surprisingly, the expected disparity was not found, and some animals even preferred stimuli presented in the previously blind hemifield. This preference persisted across multiple stimulus intensity levels and there was no indication that animals were less aware of cues in the previously blind hemifield than in its spared counterpart. Furthermore, when auditory cues were combined with visual cues, the enhanced performance they produced on a visual task was no greater in the normal than in the previously blind hemifield. These observations suggest that the multisensory rehabilitation paradigm revealed greater inherent visual information processing potential in the previously blind hemifield than was believed possible given the loss of visual cortex.     

Anatomical structures at risk using different approaches for sacrospinous ligament fixation

For 50 years now, sacrospinous ligament fixation (SSLF) has been used to treat pelvic organ prolapse consequent on altered integrity of the pelvic myofascial structures. It is usually performed vaginally, but it has recently been performed laparoscopically through either an anterior or a posterior approach, with the broad ligament as a landmark to differentiate the two. In the present study, these two laparoscopic approaches were assessed using Thiel-embalmed cadavers. The anterior and posterior approaches were compared in terms of the closest distance to anatomical structures at risk, including pelvic viscera, the obturator nerve, and vascular structures. The posterior approach was more often closer to the investigated vessels and the rectum. The obturator nerve and the ureter were close to both the anterior and posterior approaches. The urinary bladder was closer using the anterior approach. From an anatomical standpoint, therefore, the anterior laparoscopic approach for SSLF is more likely to cause injury to the urinary bladder, whereas the posterior approach is more prone to causing rectal and vessel injuries. This study illustrates, from a basic science perspective, the importance of combining fascia research, novel endoscopic or minimally invasive surgical exposures informed by anatomy, and contemporary trends in gynecology in order to improve patient outcomes. Clin. Anat. 33:522–529, 2020. © 2019 Wiley Periodicals, Inc.