Rapid, automated nerve histomorphometry through open-source artificial intelligence.

We aimed to develop and validate a deep learning model for automated segmentation and histomorphometry of myelinated peripheral nerve fibers from light microscopic images. A convolutional neural network integrated in the AxonDeepSeg framework was trained for automated axon/myelin segmentation using a dataset of light-microscopic cross-sectional images of osmium tetroxide-stained rat nerves including various axonal regeneration stages. In a second dataset, accuracy of automated segmentation was determined against manual axon/myelin labels. Automated morphometry results, including axon diameter, myelin sheath thickness and g-ratio were compared against manual straight-line measurements and morphometrics extracted from manual labels with AxonDeepSeg as a reference standard. The neural network achieved high pixel-wise accuracy for nerve fiber segmentations with a mean (± standard deviation) ground truth overlap of 0.93 (± 0.03) for axons and 0.99 (± 0.01) for myelin sheaths, respectively. Nerve fibers were identified with a sensitivity of 0.99 and a precision of 0.97. For each nerve fiber, the myelin thickness, axon diameter, g-ratio, solidity, eccentricity, orientation, and individual x -and y-coordinates were determined automatically. Compared to manual morphometry, automated histomorphometry showed superior agreement with the reference standard while reducing the analysis time to below 2.5% of the time needed for manual morphometry. This open-source convolutional neural network provides rapid and accurate morphometry of entire peripheral nerve cross-sections. Given its easy applicability, it could contribute to significant time savings in biomedical research while extracting unprecedented amounts of objective morphologic information from large image datasets.

The ultrastructural study of human cochlear nerve at different ages.

Ultrastructural and molecular changes in the myelin of the cochlear nerve (CN) have been associated with decreased hearing-acuity with increasing age. But most of these are animal studies or with very few human samples. Hence, we studied the ultrastructure of the human CN at different ages. We obtained samples of CN from persons, who at the time of death belonged to young, middle or old age-groups; defined as ≤ 30, 31 to 50, and ≥ 51 years of age, respectively. These were processed for viewing under a transmission electron microscope (TEM). Morphology and morphometry were assessed after blinding the observer. Measurements of diameter (whole nerve fibre, axon), myelin thickness and calculation of G-ratio were made on calibrated images using ImageJ software. K-Means cluster analysis was performed based on total and inner nerve fibre area. Middle and old age CN showed degenerating axons, splitting of myelin sheath and myelin balloons. Between the middle and old age groups there was significant decrease in axon diameter (p<0.001), inner nerve fibre area (p<0.001), myelin thickness (p<0.001), nerve fibre diameter (p<0.001), and G-ratio (p<0.001). By clustering, we identified three distinct populations of myelinated nerve fibres: large, medium and small. The large fibres (by size), seen in the young, disappeared in the old age-group. We were unable to find any unmyelinated nerve fibres in this study. The morphological deterioration CN fibres may be a visible sign of molecular degeneration and contribute to decreased hearing-acuity.

Human organ donor-derived vagus nerve biopsies allow for well-preserved ultrastructure and high-resolution mapping of myelinated and unmyelinated fibers.

The vagus nerve provides motor, sensory, and autonomic innervation of multiple organs, and electrical vagus nerve stimulation (VNS) provides an adjunctive treatment option for e.g. medication-refractory epilepsy and treatment-resistant depression. The mechanisms of action for VNS are not known, and high-resolution anatomical mapping of the human vagus nerve is needed to better understand its functional organization. Electron microscopy (EM) is required for the detection of both myelinated and unmyelinated axons, but access to well-preserved human vagus nerves for ultrastructural studies is sparse. Intact human vagus nerve samples were procured intra-operatively from deceased organ donors, and tissues were immediately immersion fixed and processed for EM. Ultrastructural studies of cervical and sub-diaphragmatic vagus nerve segments showed excellent preservation of the lamellated wall of myelin sheaths, and the axolemma of myelinated and unmyelinated fibers were intact. Microtubules, neurofilaments, and mitochondria were readily identified in the axoplasm, and the ultrastructural integrity of Schwann cell nuclei, Remak bundles, and basal lamina was also well preserved. Digital segmentation of myelinated and unmyelinated axons allowed for determination of fiber size and myelination. We propose a novel source of human vagus nerve tissues for detailed ultrastructural studies and mapping to support efforts to refine neuromodulation strategies, including VNS.

A shape-adjusted ellipse approach corrects for varied axonal dispersion angles and myelination in primate nerve roots.

Segmentation of axons in light and electron micrographs allows for quantitative high-resolution analysis of nervous tissues, but varied axonal dispersion angles result in over-estimates of fiber sizes. To overcome this technical challenge, we developed a novel shape-adjusted ellipse (SAE) determination of axonal size and myelination as an all-inclusive and non-biased tool to correct for oblique nerve fiber presentations. Our new resource was validated by light and electron microscopy against traditional methods of determining nerve fiber size and myelination in rhesus macaques as a model system. We performed detailed segmental mapping and characterized the morphological signatures of autonomic and motor fibers in primate lumbosacral ventral roots (VRs). An en bloc inter-subject variability for the preganglionic parasympathetic fibers within the L7-S2 VRs was determined. The SAE approach allows for morphological ground truth data collection and assignment of individual axons to functional phenotypes with direct implications for fiber mapping and neuromodulation studies.

DeepACSON automated segmentation of white matter in 3D electron microscopy.

Tracing the entirety of ultrastructures in large three-dimensional electron microscopy (3D-EM) images of the brain tissue requires automated segmentation techniques. Current segmentation techniques use deep convolutional neural networks (DCNNs) and rely on high-contrast cellular membranes and high-resolution EM volumes. On the other hand, segmenting low-resolution, large EM volumes requires methods to account for severe membrane discontinuities inescapable. Therefore, we developed DeepACSON, which performs DCNN-based semantic segmentation and shape-decomposition-based instance segmentation. DeepACSON instance segmentation uses the tubularity of myelinated axons and decomposes under-segmented myelinated axons into their constituent axons. We applied DeepACSON to ten EM volumes of rats after sham-operation or traumatic brain injury, segmenting hundreds of thousands of long-span myelinated axons, thousands of cell nuclei, and millions of mitochondria with excellent evaluation scores. DeepACSON quantified the morphology and spatial aspects of white matter ultrastructures, capturing nanoscopic morphological alterations five months after the injury.

A CNTNAP1 Missense Variant Is Associated with Canine Laryngeal Paralysis and Polyneuropathy.

Laryngeal paralysis associated with a generalized polyneuropathy (LPPN) most commonly exists in geriatric dogs from a variety of large and giant breeds. The purpose of this study was to discover the underlying genetic and molecular mechanisms in a younger-onset form of this neurodegenerative disease seen in two closely related giant dog breeds, the Leonberger and Saint Bernard. Neuropathology of an affected dog from each breed showed variable nerve fiber loss and scattered inappropriately thin myelinated fibers. Using across-breed genome-wide association, haplotype analysis, and whole-genome sequencing, we identified a missense variant in the CNTNAP1 gene (c.2810G>A; p.Gly937Glu) in which homozygotes in both studied breeds are affected. CNTNAP1 encodes a contactin-associated protein important for organization of myelinated axons. The herein described likely pathogenic CNTNAP1 variant occurs in unrelated breeds at variable frequencies. Individual homozygous mutant LPPN-affected Labrador retrievers that were on average four years younger than dogs affected by geriatric onset laryngeal paralysis polyneuropathy could be explained by this variant. Pathologic changes in a Labrador retriever nerve biopsy from a homozygous mutant dog were similar to those of the Leonberger and Saint Bernard. The impact of this variant on health in English bulldogs and Irish terriers, two breeds with higher CNTNAP1 variant allele frequencies, remains unclear. Pathogenic variants in CNTNAP1 have previously been reported in human patients with lethal congenital contracture syndrome and hypomyelinating neuropathy, including vocal cord palsy and severe respiratory distress. This is the first report of contactin-associated LPPN in dogs characterized by a deleterious variant that most likely predates modern breed establishment.

Robust Myelination of Regenerated Axons Induced by Combined Manipulations of GPR17 and Microglia.

Myelination facilitates rapid axonal conduction, enabling efficient communication across different parts of the nervous system. Here we examined mechanisms controlling myelination after injury and during axon regeneration in the central nervous system (CNS). Previously, we discovered multiple molecular pathways and strategies that could promote robust axon regrowth after optic nerve injury. However, regenerated axons remain unmyelinated, and the underlying mechanisms are elusive. In this study, we found that, in injured optic nerves, oligodendrocyte precursor cells (OPCs) undergo transient proliferation but fail to differentiate into mature myelination-competent oligodendrocytes, reminiscent of what is observed in human progressive multiple sclerosis. Mechanistically, we showed that OPC-intrinsic GPR17 signaling and sustained activation of microglia inhibit different stages of OPC differentiation. Importantly, co-manipulation of GPR17 and microglia led to extensive myelination of regenerated axons. The regulatory mechanisms of stage-dependent OPC differentiation uncovered here suggest a translatable strategy for efficient de novo myelination after CNS injury.

A computational Framework for generating rotation invariant features and its application in diffusion MRI.

In this work, we present a novel computational framework for analytically generating a complete set of algebraically independent Rotation Invariant Features (RIF) given the Laplace-series expansion of a spherical function. Our computational framework provides a closed-form solution for these new invariants, which are the natural expansion of the well known spherical mean, power-spectrum and bispectrum invariants. We highlight the maximal number of algebraically independent invariants which can be obtained from a truncated Spherical Harmonic (SH) representation of a spherical function and show that most of these new invariants can be linked to statistical and geometrical measures of spherical functions, such as the mean, the variance and the volume of the spherical signal. Moreover, we demonstrate their application to dMRI signal modeling including the Apparent Diffusion Coefficient (ADC), the diffusion signal and the fiber Orientation Distribution Function (fODF). In addition, using both synthetic and real data, we test the ability of our invariants to estimate brain tissue microstructure in healthy subjects and show that our framework provides more flexibility and open up new opportunities for innovative development in the domain of microstructure recovery from diffusion MRI.

Patterns of individual differences in fiber tract integrity of the face processing brain network support neurofunctional models.

Face cognition, the ability to perceive faces and interpret facial information, is a crucial skill in human social interactions. At the neurobiological level, several functionally specialized brain regions constitute a network of face processing. However, the evidence whether functional specialization within the face network is also reflected in the white matter structural connectivity patterns is yet limited. Based on imaging data from 1051 young healthy adult women and men, we investigated individual differences in the integrity of fibre tracts connecting face-processing regions relative to brain-general tract integrity. We analyzed individual tract-averaged fractional anisotropy (FA) values with structural equation modeling (SEM). Our results show that beyond the variance explained by a general factor indicating the quality of global tracts, the specificity of white matter integrity within the face network can be accounted for by additional factors. These factors correspond to the core and extended networks suggested in classic neuro-functional models of face processing. The right-hemisphere dominance, as commonly found in face cognition studies, is also reflected in this factorial structure. Overall, our results extend the structural brain substrate of the classic functional face processing system to the network of fibre tracts connecting these brain areas, and shed light on a structure-function correspondence from the perspective of individual differences.

Retrieving neuronal orientations using 3D scanning SAXS and comparison with diffusion MRI.

While diffusion MRI (dMRI) is currently the method of choice to non-invasively probe tissue microstructure and study structural connectivity in the brain, its spatial resolution is limited and its results need structural validation. Current ex vivo methods employed to provide 3D fiber orientations have limitations, including tissue-distorting sample preparation, small field of view or inability to quantify 3D fiber orientation distributions. 3D fiber orientation in tissue sections can be obtained from 3D scanning small-angle X-ray scattering (3D sSAXS) by analyzing the anisotropy of scattering signals. Here we adapt the 3D sSAXS method for use in brain tissue, exploiting the high sensitivity of the SAXS signal to the ordered molecular structure of myelin. We extend the characterization of anisotropy from vectors to tensors, employ the Funk-Radon-Transform for converting scattering information to real space fiber orientations, and demonstrate the feasibility of the method in thin sections of mouse brain with minimal sample preparation. We obtain a second rank tensor representing the fiber orientation distribution function (fODF) for every voxel, thereby generating fODF maps. Finally, we illustrate the potential of 3D sSAXS by comparing the result with diffusion MRI fiber orientations in the same mouse brain. We show a remarkably good correspondence, considering the orthogonality of the two methods, i.e. the different physical processes underlying the two signals. 3D sSAXS can serve as validation method for microstructural MRI, and can provide novel microstructural insights for the nervous system, given the method's orthogonality to dMRI, high sensitivity to myelin sheath's orientation and abundance, and the possibility to extract myelin-specific signal and to perform micrometer-resolution scanning.

Mouse Nr2f1 haploinsufficiency unveils new pathological mechanisms of a human optic atrophy syndrome.

Optic nerve atrophy represents the most common form of hereditary optic neuropathies leading to vision impairment. The recently described Bosch-Boonstra-Schaaf optic atrophy (BBSOA) syndrome denotes an autosomal dominant genetic form of neuropathy caused by mutations or deletions in the NR2F1 gene. Herein, we describe a mouse model recapitulating key features of BBSOA patients-optic nerve atrophy, optic disc anomalies, and visual deficits-thus representing the only available mouse model for this syndrome. Notably, Nr2f1-deficient optic nerves develop an imbalance between oligodendrocytes and astrocytes leading to postnatal hypomyelination and astrogliosis. Adult heterozygous mice display a slower optic axonal conduction velocity from the retina to high-order visual centers together with associative visual learning deficits. Importantly, some of these clinical features, such the optic nerve hypomyelination, could be rescued by chemical drug treatment in early postnatal life. Overall, our data shed new insights into the cellular mechanisms of optic nerve atrophy in BBSOA patients and open a promising avenue for future therapeutic approaches.

Time Course of Axon and Myelin Degeneration in Peripheral Nerves in Experimental Autoimmune Neuritis Rats.

Experimental autoimmune neuritis (EAN) is an animal model for Guillain-Barré syndrome (GBS), which results in neurological symptoms and histopathological changes in peripheral nerves. In this model, the correlation between the progression of the disease and the histopathological changes is not clear. To further examine histopathological changes in peripheral nerves in EAN rats, sciatic nerves were sampled at onset (day 10), peak (day 16), and recovery (days 22 and 25) of neurological symptoms in P2(57-81)-peptide-administered rats. Axon and myelin degeneration was observed by light microscopy at onset, degeneration became severe at peak, and persisted at recovery. Densities of myelinated nerve fibers and myelin areas decreased from day 10 to a minimum on day 22. Slight axon and myelin degeneration, such as accumulation of vesicles in axons and focal myelin splitting and folding, was observed by transmission electron microscopy at onset; severe degeneration, such as axonal loss, myelin ovoid, and demyelination, increased at peak; and regenerative changes, such as remyelination and enlargement of Schwann cell cytoplasm, occurred at recovery. These results suggest that EAN rats have histopathological similarities to some types of GBS patients and that EAN rats are a useful model to understand the pathogenesis of GBS.

Factors Affecting Ultrastructural Quality in the Prefrontal Cortex of the Postmortem Human Brain.

Electron microscopy (EM) studies of the postmortem human brain provide a level of resolution essential for understanding brain function in both normal and disease states. However, processes associated with death can impair the cellular and organelle ultrastructural preservation required for quantitative EM studies. Although postmortem interval (PMI), the time between death and preservation of tissue, is thought to be the most influential factor of ultrastructural quality, numerous other factors may also influence tissue preservation. The goal of the present study was to assess the effects of pre- and postmortem factors on multiple components of ultrastructure in the postmortem human prefrontal cortex. Tissue samples from 30 subjects were processed using standard EM histochemistry. The primary dependent measure was number of identifiable neuronal profiles, and secondary measures included presence and/or integrity of synapses, mitochondria, and myelinated axonal fibers. Number of identifiable neuronal profiles was most strongly affected by the interaction of PMI and pH, such that short PMIs and neutral pH values predicted the best preservation. Secondary measures were largely unaffected by pre- and postmortem factors. Together, these data indicate that distinct components of the neuropil are differentially affected by PMI and pH in postmortem human brain.

Morphometric analysis of peripheral myelinated nerve fibers through deep learning.

Irrespective of initial causes of neurological diseases, these disorders usually exhibit two key pathological changes-axonal loss or demyelination or a mixture of the two. Therefore, vigorous quantification of myelin and axons is essential in studying these diseases. However, the process of quantification has been labor intensive and time-consuming because of the requisite manual segmentation of myelin and axons from microscopic nerve images. As a part of AI development, deep learning has been utilized to automate certain tasks, such as image analysis. This study describes the development of a convolutional neural network (CNN)-based approach to segment images of mouse nerve cross sections. We adapted the U-Net architecture and used manually-produced segmentation data accumulated over many years in our lab for training. These images ranged from normal nerves to those afflicted by severe myelin and axon pathologies; thus, maximizing the trained model's ability to recognize atypical myelin structures. Morphometric data produced by applying the trained model to additional images were then compared to manually obtained morphometrics. The former effectively shortened the time consumption in the morphometric analysis with excellent accuracy in axonal density and g-ratio. However, we were not able to completely eliminate manual refinement of the segmentation product. We also observed small variations in axon diameter and myelin thickness within 9.5%. Nevertheless, we learned alternative ways to improve accuracy through the study. Overall, greatly increased efficiency in the CNN-based approach out-weighs minor limitations that will be addressed in future studies, thus justifying our confidence in its prospects. Note: All the relevant code is freely available at https://neurology.med.wayne.edu/drli-datashairing.

Carbofuran hampers oligodendrocytes development leading to impaired myelination in the hippocampus of rat brain.

During the mammalian brain development, oligodendrocyte progenitor cells (OPCs) are generated from neuroepithelium and migrate throughout the brain. Myelination is a tightly regulated process which involves time framed sequential events of OPCs proliferation, migration, differentiation and interaction with axons for functional insulated sheath formation. Myelin is essential for efficient and rapid conduction of electric impulses and its loss in the hippocampus of the brain may result in impaired memory and long-term neurological deficits. Carbofuran, a carbamate pesticide is known to cause inhibition of hippocampal neurogenesis and memory dysfunctions in rats. Nonetheless, the effects of carbofuran on OPCs proliferation, fate determination, maturation/differentiation and myelination potential in the hippocampus of the rat brain are still completely elusive. Herein, we investigated the effects of sub-chronic exposure of carbofuran during two different time periods including prenatal and adult brain development in rats. We observed carbofuran hampers OPCs proliferation (BrdU incorporation) and oligodendroglial differentiation in vitro. Similar effects of carbofuran were also observed in the hippocampus region of the brain at both the time points. Carbofuran exposure resulted in reduced expression of key genes and proteins involved in the regulation of oligodendrocyte development and functional myelination. It also affects the survival of oligodendrocytes by inducing apoptotic cell death. The ultrastructural analysis of myelin architecture clearly depicted carbofuran-mediated negative effects on myelin compaction and g-ratio alteration. Conclusively, our study demonstrated that carbofuran alters myelination potential in the hippocampus, which leads to cognitive deficits in rats.

Changes in white matter and the effects of fluoxetine on such changes in the CUS rat model of depression.

OBJECTIVE: To explore the pathogenesis of depression and the possible mechanism of the effects of selective serotonin reuptake inhibitors (SSRIs) on the myelinated fibers and myelin sheaths in the white matter during the antidepressant action of fluoxetine. METHODS: In this study, Sprague Dawley (SD) rats were divided into a Control group, a group treated with CUS and no drugs (CUS/Standard group) and a group treated with CUS and fluoxetine (CUS/FLX group). The CUS/FLX group was treated with fluoxetine at dose of 5 mg/kg for 21 days. The white matter volume, the myelinated fiber parameters and the myelin sheath volume in the white matter were calculated from transmission electron microscope images through unbiased stereological methods. RESULTS: The total volume and total length of myelinated fibers;and mean volume of white matter of the CUS/Standard group were significantly decreased compared to values from the control group (p = 0.025, p = 0.007, p = 0.000), whereas no significant differences in these stereological parameters were found between the CUS/Standard and CUS/FLX groups (p > 0.05). CONCLUSIONS: Fluoxetine successfully treated depression-like behavior but had no effects on the white matter or its component myelinated fibers in the CUS rat model of depression.

The leptin receptor mutation of the obese Zucker rat causes sciatic nerve demyelination with a centripetal pattern defect.

Young male Zucker rats with a leptin receptor mutation are obese, have a non-insulin-dependent diabetes mellitus (NIDDM), and other endocrinopathies. Tibial branches of the sciatic nerve reveal a progressive demyelination that progresses out of the Schwann cells (SCs) where electron-contrast deposits are accumulated while the minor lines or intermembranous SC contacts display exaggerated spacings. Cajal bands contain diversely contrasted vesicles adjacent to the abaxonal myelin layer with blemishes; they appear dispatched centripetally out of many narrow electron densities, regularly spaced around the myelin annulus. These anomalies widen and yield into sectors across the stacked myelin layers. Throughout the worse degradations, the adaxonal membrane remains along the axonal neuroplasm. This peripheral neuropathy with irresponsive leptin cannot modulate hypothalamic-pituitary-adrenal axis and SC neurosteroids, thus exacerbates NIDDM condition. Additionally, the ultrastructure of the progressive myelin alterations may have unraveled a peculiar, centripetal mode of trafficking maintenance of the peripheral nervous system myelin, while some adhesive glycoproteins remain between myelin layers, somewhat hindering the axon mutilation. Heading title: Peripheral neuropathy and myelin.

A novel cerebellar commissure and other myelinated axons in the Purkinje cell layer of a pond turtle (Trachemys scripta elegans).

Parallel fibers in the molecular layer of the vertebrate cerebellum mediate slow spike conduction in the transverse plane. In contrast, electrophysiological recordings have indicated that rapid spike conduction exists between the lateral regions of the cerebellar cortex of the red-ear pond turtle (Trachemys scripta). The anatomical basis for this commissure is now examined in that species using neuronal tracing techniques. Fluorescently tagged dextrans and lipophilic carbocyanine dyes placed in one lateral edge of this nonfoliated cortex are transported across the midline of living brains in vitro and along the axonal membranes of fixed tissues, respectively. Surprisingly, the labeled commissural axons traversed the cortex within the Purkinje cell layer, and not in the white matter of the molecular layer or the white matter below the granule cell layer. Unlike thin parallel fibers that exhibit characteristic varicosities, this commissure is composed of smooth axons of large diameter that also extend beyond the cerebellar cortex via the cerebellar peduncles. Double labeling with myelin basic protein antibody demonstrated that these commissural axons are ensheathed with myelin. In contrast to this transverse pathway, an orthogonal myelinated tract was observed along the cerebellar midline. The connections of this transverse commissure with the lateral cerebellum, the vestibular nuclear complex, and the cochlear vestibular ganglia indicate that this commissure plays a role in bilateral vestibular connectivity.

3D microsurgical anatomy of the cortico-spinal tract and lemniscal pathway based on fiber microdissection and demonstration with tractography. / Anatomía microquirúrgica en 3D del tracto corticoespinal y de la vía del lemnisco basada en microdisección de fibras y demostración a través de tractografía.

OBJECTIVE: To demonstrate tridimensionally the anatomy of the cortico-spinal tract and the medial lemniscus, based on fiber microdissection and diffusion tensor tractography (DTT). MATERIAL AND METHODS: Ten brain hemispheres and brain-stem human specimens were dissected and studied under the operating microscope with microsurgical instruments by applying the fiber microdissection technique. Brain magnetic resonance imaging was obtained from 15 healthy subjects using diffusion-weighted images, in order to reproduce the cortico-spinal tract and the lemniscal pathway on DTT images. RESULTS: The main bundles of the cortico-spinal tract and medial lemniscus were demonstrated and delineated throughout most of their trajectories, noticing their gross anatomical relation to one another and with other white matter tracts and gray matter nuclei the surround them, specially in the brain-stem; together with their corresponding representation on DTT images. CONCLUSIONS: Using the fiber microdissection technique we were able to distinguish the disposition, architecture and general topography of the cortico-spinal tract and medial lemniscus. This knowledge has provided a unique and profound anatomical perspective, supporting the correct representation and interpretation of DTT images. This information should be incorporated in the clinical scenario in order to assist surgeons in the detailed and critic analysis of lesions located inside the brain-stem, and therefore, improve the surgical indications and planning, including the preoperative selection of optimal surgical strategies and possible corridors to enter the brainstem, to achieve safer and more precise microsurgical technique.