AxonSeg: Open Source Software for Axon and Myelin Segmentation and Morphometric Analysis

Developmental origin of oligodendrocytes determines their function in the adult brain

In the mouse embryonic forebrain, developmentally distinct oligodendrocyte progenitor cell populations and their progeny, oligodendrocytes, emerge from three distinct regions in a spatiotemporal gradient from ventral to dorsal. However, the functional importance of this oligodendrocyte developmental heterogeneity is unknown. Using a genetic strategy to ablate dorsally derived oligodendrocyte lineage cells (OLCs), we show here that the areas in which dorsally derived OLCs normally reside in the adult central nervous system become populated and myelinated by OLCs of ventral origin. These ectopic oligodendrocytes (eOLs) have a distinctive gene expression profile as well as subtle myelination abnormalities. The failure of eOLs to fully assume the role of the original dorsally derived cells results in locomotor and cognitive deficits in the adult animal. This study reveals the importance of developmental heterogeneity within the oligodendrocyte lineage and its importance for homeostatic brain function.

Vibrational spectroscopy and multiphoton microscopy for label-free visualization of nervous system degeneration and regeneration

Neurological disorders, including spinal cord injury, peripheral nerve injury, traumatic brain injury, and neurodegenerative diseases, pose significant challenges in terms of diagnosis, treatment, and understanding the underlying pathophysiological processes. Label-free multiphoton microscopy techniques, such as coherent Raman scattering, two-photon excited autofluorescence, and second and third harmonic generation microscopy, have emerged as powerful tools for visualizing nervous tissue with high resolution and without the need for exogenous labels. Coherent Raman scattering processes as well as third harmonic generation enable label-free visualization of myelin sheaths, while their combination with two-photon excited autofluorescence and second harmonic generation allows for a more comprehensive tissue visualization. They have shown promise in assessing the efficacy of therapeutic interventions and may have future applications in clinical diagnostics. In addition to multiphoton microscopy, vibrational spectroscopy methods such as infrared and Raman spectroscopy offer insights into the molecular signatures of injured nervous tissues and hold potential as diagnostic markers. This review summarizes the application of these label-free optical techniques in preclinical models and illustrates their potential in the diagnosis and treatment of neurological disorders with a special focus on injury, degeneration, and regeneration. Furthermore, it addresses current advancements and challenges for bridging the gap between research findings and their practical applications in a clinical setting.

NRV: An open framework for in silico evaluation of peripheral nerve electrical stimulation strategies

Electrical stimulation of peripheral nerves has been used in various pathological contexts for rehabilitation purposes or to alleviate the symptoms of neuropathologies, thus improving the overall quality of life of patients. However, the development of novel therapeutic strategies is still a challenging issue requiring extensive in vivo experimental campaigns and technical development. To facilitate the design of new stimulation strategies, we provide a fully open source and self-contained software framework for the in silico evaluation of peripheral nerve electrical stimulation. Our modeling approach, developed in the popular and well-established Python language, uses an object-oriented paradigm to map the physiological and electrical context. The framework is designed to facilitate multi-scale analysis, from single fiber stimulation to whole multifascicular nerves. It also allows the simulation of complex strategies such as multiple electrode combinations and waveforms ranging from conventional biphasic pulses to more complex modulated kHz stimuli. In addition, we provide automated support for stimulation strategy optimization and handle the computational backend transparently to the user. Our framework has been extensively tested and validated with several existing results in the literature.

CellWalker: a user-friendly and modular computational pipeline for morphological analysis of microscopy images

SUMMARY

The implementation of computational tools for analysis of microscopy images has been one of the most important technological innovations in biology, providing researchers unmatched capabilities to comprehend cell shape and connectivity. While numerous tools exist for image annotation and segmentation, there is a noticeable gap when it comes to morphometric analysis of microscopy images. Most existing tools often measure features solely on 2D serial images, which can be difficult to extrapolate to 3D. For this reason, we introduce CellWalker, a computational toolbox that runs inside Blender, an open-source computer graphics software. This add-on improves the morphological analysis by seamlessly integrating analysis tools into the Blender workflow, providing visual feedback through a powerful 3D visualization, and leveraging the resources of Blender's community. CellWalker provides several morphometric analysis tools that can be used to calculate distances, volume, surface areas and to determine cross-sectional properties. It also includes tools to build skeletons, calculate distributions of subcellular organelles. In addition, this python-based tool contains 'visible-source' IPython notebooks accessories for segmentation of 2D/3D microscopy images using deep learning and visualization of the segmented images that are required as input to CellWalker. Overall, CellWalker provides practical tools for segmentation and morphological analysis of microscopy images in the form of an open-source and modular pipeline which allows a complete access to fine-tuning of algorithms through visible-source code while still retaining a result-oriented interface.

AVAILABILITY AND IMPLEMENTATION

CellWalker source code is available on GitHub (https://github.com/utraf-pasteur-institute/Cellwalker-blender and https://github.com/utraf-pasteur-institute/Cellwalker-notebooks) under a GPL-3 license.

AimSeg: A machine-learning-aided tool for axon, inner tongue and myelin segmentation

Electron microscopy (EM) images of axons and their ensheathing myelin from both the central and peripheral nervous system are used for assessing myelin formation, degeneration (demyelination) and regeneration (remyelination). The g-ratio is the gold standard measure of assessing myelin thickness and quality, and traditionally is determined from measurements made manually from EM images-a time-consuming endeavour with limited reproducibility. These measurements have also historically neglected the innermost uncompacted myelin sheath, known as the inner tongue. Nonetheless, the inner tongue has been shown to be important for myelin growth and some studies have reported that certain conditions can elicit its enlargement. Ignoring this fact may bias the standard g-ratio analysis, whereas quantifying the uncompacted myelin has the potential to provide novel insights in the myelin field. In this regard, we have developed AimSeg, a bioimage analysis tool for axon, inner tongue and myelin segmentation. Aided by machine learning classifiers trained on transmission EM (TEM) images of tissue undergoing remyelination, AimSeg can be used either as an automated workflow or as a user-assisted segmentation tool. Validation results on TEM data from both healthy and remyelinating samples show good performance in segmenting all three fibre components, with the assisted segmentation showing the potential for further improvement with minimal user intervention. This results in a considerable reduction in time for analysis compared with manual annotation. AimSeg could also be used to build larger, high quality ground truth datasets to train novel deep learning models. Implemented in Fiji, AimSeg can use machine learning classifiers trained in ilastik. This, combined with a user-friendly interface and the ability to quantify uncompacted myelin, makes AimSeg a unique tool to assess myelin growth.

TREM2-dependent microglial function is essential for remyelination and subsequent neuroprotection

Disability in multiple sclerosis (MS) is driven in part by the failure of remyelination and progressive neurodegeneration. Microglia, and specifically triggering receptor expressed on myeloid cells 2 (TREM2), a factor highly expressed in microglia, have been shown to play an important role in remyelination. Here, using a focal demyelination model in the brain, we demonstrate that demyelination is persistent in TREM2 knockout mice, lasting more than 6 weeks after lysolecithin injection and resulting in substantial neurodegeneration. We also find that TREM2 knockout mice exhibit an altered glial response following demyelination. TREM2 knockout microglia demonstrate defects in migration and phagocytosis of myelin debris. In addition, human monocyte-derived macrophages from subjects with a TREM2 mutation prevalent in human disease also show a defect in myelin debris phagocytosis. Together, we highlight the central role of TREM2 signaling in remyelination and neuroprotection. These findings provide insights into how chronic demyelination might lead to axonal damage and could help identify novel neuroprotective therapeutic targets for MS.

AxoNet 2.0: A Deep Learning-Based Tool for Morphometric Analysis of Retinal Ganglion Cell Axons

Purpose

Assessment of glaucomatous damage in animal models is facilitated by rapid and accurate quantification of retinal ganglion cell (RGC) axonal loss and morphologic change. However, manual assessment is extremely time- and labor-intensive. Here, we developed AxoNet 2.0, an automated deep learning (DL) tool that (i) counts normal-appearing RGC axons and (ii) quantifies their morphometry from light micrographs.

Methods

A DL algorithm was trained to segment the axoplasm and myelin sheath of normal-appearing axons using manually-annotated rat optic nerve (ON) cross-sectional micrographs. Performance was quantified by various metrics (e.g., soft-Dice coefficient between predicted and ground-truth segmentations). We also quantified axon counts, axon density, and axon size distributions between hypertensive and control eyes and compared to literature reports.

Results

AxoNet 2.0 performed very well when compared to manual annotations of rat ON (R2 = 0.92 for automated vs. manual counts, soft-Dice coefficient = 0.81 ± 0.02, mean absolute percentage error in axonal morphometric outcomes < 15%). AxoNet 2.0 also showed promise for generalization, performing well on other animal models (R2 = 0.97 between automated versus manual counts for mice and 0.98 for non-human primates). As expected, the algorithm detected decreased in axon density in hypertensive rat eyes (P ≪ 0.001) with preferential loss of large axons (P < 0.001).

Conclusions

AxoNet 2.0 provides a fast and nonsubjective tool to quantify both RGC axon counts and morphological features, thus assisting with assessing axonal damage in animal models of glaucomatous optic neuropathy.

Translational Relevance

This deep learning approach will increase rigor of basic science studies designed to investigate RGC axon protection and regeneration.

Histology-informed automatic parcellation of white matter tracts in the rat spinal cord

The white matter is organized into “tracts” or “bundles,” which connect different parts of the central nervous system. Knowing where these tracts are located in each individual is important for understanding the cause of potential sensorial, motor or cognitive deficits and for developing appropriate treatments. Traditionally, tracts are found using tracer injection, which is a difficult, slow and poorly scalable technique. However, axon populations from a given tract exhibit specific characteristics in terms of morphometrics and myelination. Hence, the delineation of tracts could, in principle, be done based on their morphometry. The objective of this study was to generate automatic parcellation of the rat spinal white matter tracts using the manifold information from scanning electron microscopy images of the entire spinal cord. The axon morphometrics (axon density, axon diameter, myelin thickness and g-ratio) were computed pixelwise following automatic axon segmentation using AxonSeg. The parcellation was based on an agglomerative clustering algorithm to group the tracts. Results show that axon morphometrics provide sufficient information to automatically identify some white matter tracts in the spinal cord, however, not all tracts were correctly identified. Future developments of microstructure quantitative MRI even bring hope for a personalized clustering of white matter tracts in each individual patient. The generated atlas and the associated code can be found at https://github.com/neuropoly/tract-clustering.

Protocol to image and analyze the morphology of mouse peripheral nerves using transmission electron microscopy

Morphological analysis of peripheral nerves in mouse models can be used to characterize the pathophysiology of peripheral nerve disease, but obtaining high-quality electron micrographs can be challenging. Here, we present a protocol to obtain electron micrographs of mouse peripheral nerves. We detail the procedures of sampling, fixation, and embedding of peripheral nerves. We then outline the steps for ultrathin sectioning and transmission electron microscopy imaging. Finally, we describe morphological evaluation of nerve fibers in these images using ImageJ and AxonSeg.

For complete details on the use and execution of this protocol, please refer to Nakai-Shimoda et al. (2021).

•

Morphological analysis of mouse peripheral nerves using transmission electron microscopy

•

Step-by-step guide for sampling, fixation, and embedding of peripheral nerves

•

Ultrathin sectioning by glass knife and imaging using transmission electron microscopy

•

Image analysis of unmyelinated and myelinated nerve fibers using ImageJ and AxonSeg

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Role of Dectin-1 in peripheral nerve injury

Dectin-1, a C-type lectin receptor, plays a role in nerve injury in the central nervous system. However, whether it plays a role in the peripheral nervous system is not well understood. Our study showed the expression of Dectin-1 on the membrane of macrophages. We also used a sciatic nerve crushing injury model to demonstrate that there was a delay in nerve degeneration-related processes such as breakdown of injured myelinated nerve fibers and formation of myelin ovoid in groups injected with whole glucan particle soluble (WGPS), a Dectin-1 antagonist. There were also fewer intraneural blood vessels in the Dectin-1 antagonist treated group. Our study suggested inhibiting Dectin-1 delayed debris clearance, nerve degeneration, and angiogenesis after peripheral nerve injury.

Enhanced Automatic Morphometry of Nerve Histological Sections Using Ensemble Learning

There is a need for an automated morphometry algorithm to facilitate the otherwise labor-intensive task of the quantitative histological analysis of neural microscopic images. A benchmark morphometry algorithm is the convolutional neural network Axondeepseg (ADS), which yields a high segmentation accuracy for scanning and transmission electron microscopy images. Nevertheless, it shows decreased accuracy when applied to optical microscopy images, and it has been observed to yield sizable false positives when identifying small-sized neurons within the slides. In this study, ensemble learning is used to enhance the performance of ADS by combining it with the paired image-to-image translation algorithm PairedImageTranslation (PIT). Here, 120 optical microscopy images of peripheral nerves were used to train and test the ensemble learning model and the two base models individually for comparison. The results showed weighted pixel-wise accuracy for the ensemble model of 95.5%, whereas the ADS and PIT yielded accuracies of 93.4% and 90%, respectively. The automated measurements of the axon diameters and myelin thicknesses from the manually marked ground truth images were not statistically different (p = 0.05) from the measurements taken from the same images when segmented using the developed ensemble model, while they were different when they were measured from the segmented images by the two base models individually. The automated measurement of the G ratios indicated a higher similarity to the ground truth testing images for the ensemble model in comparison with the individual base models. The proposed model yielded automated segmentation of the nerve slides, which were sufficiently equivalent to the manual annotations and could be employed for axon diameters and myelin thickness measurements for fully automated histological analysis of the neural images.

High-throughput segmentation of unmyelinated axons by deep learning

Axonal characterizations of connectomes in healthy and disease phenotypes are surprisingly incomplete and biased because unmyelinated axons, the most prevalent type of fibers in the nervous system, have largely been ignored as their quantitative assessment quickly becomes unmanageable as the number of axons increases. Herein, we introduce the first prototype of a high-throughput processing pipeline for automated segmentation of unmyelinated fibers. Our team has used transmission electron microscopy images of vagus and pelvic nerves in rats. All unmyelinated axons in these images are individually annotated and used as labeled data to train and validate a deep instance segmentation network. We investigate the effect of different training strategies on the overall segmentation accuracy of the network. We extensively validate the segmentation algorithm as a stand-alone segmentation tool as well as in an expert-in-the-loop hybrid segmentation setting with preliminary, albeit remarkably encouraging results. Our algorithm achieves an instance-level [Formula: see text] score of between 0.7 and 0.9 on various test images in the stand-alone mode and reduces expert annotation labor by 80% in the hybrid setting. We hope that this new high-throughput segmentation pipeline will enable quick and accurate characterization of unmyelinated fibers at scale and become instrumental in significantly advancing our understanding of connectomes in both the peripheral and the central nervous systems.

MyelTracer: A Semi-Automated Software for Myelin g-Ratio Quantification

In the central and peripheral nervous systems, the myelin sheath promotes neuronal signal transduction. The thickness of the myelin sheath changes during development and in disease conditions like multiple sclerosis. Such changes are routinely detected using electron microscopy through g-ratio quantification. While g-ratio is one of the most critical measurements in myelin studies, a major drawback is that g-ratio quantification is extremely laborious and time-consuming. Here, we report the development and validation of MyelTracer, an installable, stand-alone software for semi-automated g-ratio quantification based on the Open Computer Vision Library (OpenCV). Compared with manual g-ratio quantification, using MyelTracer produces consistent results across multiple tissues and animal ages, as well as in remyelination after optic nerve crush, and reduces total quantification time by 40-60%. With g-ratio measurements via MyelTracer, a known hypomyelination phenotype can be detected in a Williams syndrome mouse model. MyelTracer is easy to use and freely available for Windows and Mac OS X (https://github.com/HarrisonAllen/MyelTracer).

Decoding the microstructural properties of white matter using realistic models

Multi-echo gradient echo (ME-GRE) magnetic resonance signal evolution in white matter has a strong dependence on the orientation of myelinated axons with respect to the main static field. Although analytical solutions have been able to predict some of the white matter (WM) signal behaviour of the hollow cylinder model, it has been shown that realistic models of WM offer a better description of the signal behaviour observed. In this work, we present a pipeline to (i) generate realistic 2D WM models with their microstructure based on real axon morphology with adjustable fiber volume fraction (FVF) and g-ratio. We (ii) simulate their interaction with the static magnetic field to be able to simulate their MR signal. For the first time, we (iii) demonstrate that realistic 2D WM models can be used to simulate a MR signal that provides a good approximation of the signal obtained from a real 3D WM model derived from electron microscopy. We then (iv) demonstrate in silico that 2D WM models can be used to predict microstructural parameters in a robust way if ME-GRE multi-orientation data is available and the main fiber orientation in each pixel is known using DTI. A deep learning network was trained and characterized in its ability to recover the desired microstructural parameters such as FVF, g-ratio, free and bound water transverse relaxation and magnetic susceptibility. Finally, the network was trained to recover these micro-structural parameters from an ex vivo dataset acquired in 9 orientations with respect to the magnetic field and 12 echo times. We demonstrate that this is an overdetermined problem and that as few as 3 orientations can already provide comparable results for some of the decoded metrics.

Can MRI measure myelin? Systematic review, qualitative assessment, and meta-analysis of studies validating microstructural imaging with myelin histology

Recent years have seen an increased understanding of the importance of myelination in healthy brain function and neuropsychiatric diseases. Non-invasive microstructural magnetic resonance imaging (MRI) holds the potential to expand and translate these insights to basic and clinical human research, but the sensitivity and specificity of different MR markers to myelination is a subject of debate. To consolidate current knowledge on the topic, we perform a systematic review and meta-analysis of studies that validate microstructural imaging by combining it with myelin histology. We find meta-analytic evidence for correlations between various myelin histology metrics and markers from different MRI modalities, including fractional anisotropy, radial diffusivity, macromolecular pool, magnetization transfer ratio, susceptibility and longitudinal relaxation rate, but not mean diffusivity. Meta-analytic correlation effect sizes range widely, between R2 = 0.26 and R2 = 0.82. However, formal comparisons between MRI-based myelin markers are limited by methodological variability, inconsistent reporting and potential for publication bias, thus preventing the establishment of a single most sensitive strategy to measure myelin with MRI. To facilitate further progress, we provide a detailed characterisation of the evaluated studies as an online resource. We also share a set of 12 recommendations for future studies validating putative MR-based myelin markers and deploying them in vivo in humans.

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.

Myelin detection in fluorescence microscopy images using machine learning

BACKGROUND

The myelin sheath produced by glial cells insulates the axons, and supports the function of the nervous system. Myelin sheath degeneration causes neurodegenerative disorders, such as multiple sclerosis (MS). There are no therapies for MS that promote remyelination. Drug discovery frequently involves screening thousands of compounds. However, this is not feasible for remyelination drugs, since myelin quantification is a manual labor-intensive endeavor. Therefore, the development of assistive software for expedited myelin detection is instrumental for MS drug discovery by enabling high-content image-based drug screens.

NEW METHOD

In this study, we developed a machine learning based expedited myelin detection approach in fluorescence microscopy images. Multi-channel three-dimensional microscopy images of a mouse stem cell-based myelination assay were labeled by experts. A spectro-spatial feature extraction method was introduced to represent local dependencies of voxels both in spatial and spectral domains. Feature extraction yielded two data set of over forty-seven thousand annotated images in total.

RESULTS

Myelin detection performances of 23 different supervised machine learning techniques including a customized-convolutional neural network (CNN), were assessed using various train/test split ratios of the data sets. The highest accuracy values of 98.84±0.09% and 98.46±0.11% were achieved by Boosted Trees and customized-CNN, respectively.

COMPARISON WITH EXISTING METHODS

Our approach can detect myelin in a common experimental setup. Myelin extending in any orientation in 3 dimensions is segmented from 3 channel z-stack fluorescence images.

CONCLUSIONS

Our results suggest that the proposed expedited myelin detection approach is a feasible and robust method for remyelination drug screening.

Dendritic Degeneration of Human Auditory Nerve Fibers and Its Impact on the Spiking Pattern Under Regular Conditions and During Cochlear Implant Stimulation

Due to limitations of human in vivo studies, detailed computational models enable understanding the neural signaling in the degenerated auditory system and cochlear implants (CIs). Four human cochleae were used to quantify hearing levels depending on dendritic changes in diameter and myelination thickness from type I of the auditory nerve fibers (ANFs). Type I neurons transmit the auditory information as spiking pattern from the inner hair cells (IHCs) to the cochlear nucleus. The impact of dendrite diameter and degree of myelination on neural signal transmission was simulated for (1) synaptic excitation via IHCs and (2) stimulation from CI electrodes. An accurate three-dimensional human cochlear geometry, along with 30 auditory pathways, mimicked the CI environment. The excitation properties of electrical potential distribution induced by two CI were analyzed. Main findings: (1) The unimodal distribution of control dendrite diameters becomes multimodal for hearing loss cases; a group of thin dendrites with diameters between 0.3 and 1 μm with a peak at 0.5 μm appeared. (2) Postsynaptic currents from IHCs excite such thin dendrites easier and earlier than under control conditions. However, this advantage is lost as their conduction velocity decreases proportionally with the diameter and causes increased spike latency and jitter in soma and axon. Firing probability reduces through the soma passage due to the low intracellular current flow in thin dendrites during spiking. (3) Compared with dendrite diameter, variations in myelin thickness have a small impact on spiking performance. (4) Contrary to synaptic excitation, CIs cause several spike initiation sites in dendrite, soma region, and axon; moreover, fiber excitability reduces with fiber diameter. In a few cases, where weak stimuli elicit spikes of a target neuron (TN) in the axon, dendrite diameter reduction has no effect. However, in many cases, a spike in a TN is first initiated in the dendrite, and consequently, dendrite degeneration demands an increase in threshold currents. (5) Threshold currents of a TN and co-stimulation of degenerated ANFs in other frequency regions depend on the electrode position, including its distance to the outer wall, the cochlear turn, and the three-dimensional pathway of the TN.

Atypical myelinogenesis and reduced axon caliber in the Scn1a variant model of Dravet syndrome: An electron microscopy pilot study of the developing and mature mouse corpus callosum

Dravet Syndrome (DS) is a genetic neurodevelopmental disease. Recurrent severe seizures begin in infancy and co-morbidities follow, including developmental delay, cognitive and behavioral dysfunction. A majority of DS patients have an SCN1A heterozygous gene mutation. This mutation causes a loss-of-function in inhibitory neurons, initiating seizure onset. We have investigated whether the sodium channelopathy may result in structural changes in the DS model independent of seizures. Morphometric analyses of axons within the corpus callosum were completed at P16 and P50 in Scn1a heterozygote KO male mice and their age-matched wild-type littermates. Trainable machine learning algorithms were used to examine electron microscopy images of ~400 myelinated axons per animal, per genotype, including myelinated axon cross-section area, frequency distribution and g-ratios. Pilot data for Scn1a heterozygote KO mice demonstrate the average axon caliber was reduced in developing and adult mice. Qualitative analysis also shows micro-features marking altered myelination at P16 in the DS model, with myelin out-folding and myelin debris within phagocytic cells. The data has indicated, in the absence of behavioral seizures, factors that governed a shift toward small calibre axons at P16 have persisted in adult Scn1a heterozygote KO corpus callosum. The pilot study provides a basis for future meta-analysis that will enable robust estimates of the effects of the sodium channelopathy on axon architecture. We propose that early therapeutic strategies in DS could help minimize the effect of sodium channelopathies, beyond the impact of overt seizures, and therefore achieve better long-term treatment outcomes.

Preventive hypothermia as a neuroprotective strategy for paclitaxel-induced peripheral neuropathy

Chemotherapy-induced peripheral neuropathy (CIPN) is a severe adverse effect that occurs secondary to anticancer treatments and has no known preventive or therapeutic strategy. Therapeutic hypothermia has been shown to be effective in protecting against central and peripheral nervous system injuries. However, the effects of therapeutic hypothermia on CIPN have rarely been explored. We induced lower back hypothermia (LBH) in an established paclitaxel-induced CIPN rat model and found that the paclitaxel-induced impairments observed in behavioral, electrophysiological, and histological impairments were inhibited by LBH when applied at an optimal setting of 24°C to the sciatic nerve and initiated 90 minutes before paclitaxel infusion. Lower back hypothermia also inhibited the paclitaxel-induced activation of astroglia and microglia in the spinal cord and macrophage infiltration into and neuronal injury in the dorsal root ganglia and sciatic nerves. Furthermore, LBH decreased the local blood flow and local tissue concentrations of paclitaxel. Finally, in NOD/SCID mice inoculated with cancer cells, the antiproliferative effect of paclitaxel was not affected by the distal application of LBH. In conclusion, our findings indicate that early exposure to regional hypothermia alleviates paclitaxel-induced peripheral neuropathy. Therapeutic hypothermia may therefore represent an economical and nonpharmaceutical preventive strategy for CIPN in patients with localized solid tumors.

Blockade of IL-6 signaling prevents paclitaxel-induced neuropathy in C57Bl/6 mice

The microtubule-stabilizing agent paclitaxel frequently leads to chemotherapy-induced peripheral neuropathy (CIN), which further increases the burden of disease and often necessitates treatment limitations. The pathophysiology of CIN appears to involve both “upstream” effects including altered intracellular calcium signaling and activation of calcium dependent proteases such as calpain as well as subsequent “downstream” neuro-inflammatory reactions with cytokine release and macrophage infiltration of dorsal root ganglia. In this study, we aimed to investigate whether these processes are linked by the pro-inflammatory cytokine interleukin-6 (IL-6). We observed that paclitaxel exposure induced IL-6 synthesis in cultured sensory neurons from postnatal Wistar rats, which could be prevented by co-treatment with a calpain inhibitor. This suggests a calcium dependent process. We demonstrate that adult C57BL/6 mice deficient in IL-6 are protected from developing functional and histological changes of paclitaxel-induced neuropathy. Furthermore, pretreatment with an IL-6-neutralizing antibody resulted in the prevention of paclitaxel-induced neuropathy in C57BL/6 mice. Electrophysiological data from our preclinical model was adequately reflected by measurements of patients undergoing paclitaxel therapy for ovarian cancer. In this cohort, measured Il-6 levels correlated with the severity of neuropathy. Our findings demonstrate that IL-6 plays a pivotal role in the pathophysiology of paclitaxel-induced neuropathy per se and that pharmacological or genetic interference with this signaling pathway prevents the development of this potentially debilitating adverse effect. These findings provide a rationale for a clinical trial with IL-6 neutralizing antibodies to prevent dose-limiting neurotoxic adverse effects of paclitaxel chemotherapy.

What Has Been Learned from Magnetic Resonance Imaging Examination of the Injured Human Spinal Cord: A Canadian Perspective

Magnetic resonance imaging (MRI) has transformed the way surgeons and researchers study and treat spinal cord injury. In this narrative review, we explore the historical context of imaging the human spinal cord and describe how MRI has evolved from providing the first visualization of the human spinal cord in the 1980s to a remarkable set of imaging tools today. The article focuses in particular on the role of Canadian researchers to this field. We begin by outlining the clinical context of traumatic injury to the human spinal cord and describe why current MRI standards fall short when it comes to treating this disabling condition. Parts 2 and 3 of this work explore an exciting and dramatic shift in the use of MRI technology to aid in our understanding and treatment of traumatic injury to the spinal cord. We explore the use of functional imaging (part 2) and structural imaging (part 3) and explore how these techniques have evolved, how they are used, and the challenges that we face for continued refinement and application to patients who live with the neurological and functional deficits caused by injury to the delicate spinal cord.

Histological and diffusion-weighted magnetic resonance imaging data from normal and degenerated optic nerve and chiasm of the rat

Diffusion-weighted magnetic resonance imaging (dMRI) is widely used to infer microstructural characteristics of tissue, particularly in cerebral white matter. Histological validation of the metrics derived from dMRI methods are needed to fully characterize their ability to capture biologically-relevant histological features non-invasively. The data described here were used to correlate metrics derived from dMRI and quantitative histology in an animal model of axonal degeneration (“Histological validation of per-bundle water diffusion metrics within a region of fiber crossing following axonal degeneration” [1]). Unilateral retinal ischemia/reperfusion was induced in 10 rats, by the elevation of pressure of the anterior chamber of the eye for 90 min. Five rats were used as controls. After five weeks, injured animals were intracardially perfused to analyze the optic nerves and chiasm with dMRI and histology. This resulted in 15 brain scans, each with 80 diffusion-sensitizing gradient directions with b = 2000 and 2500 s/mm² and 20 non-diffusion-weighted images (b = 0 s/mm²), with isometric voxel resolution of 125 μm³. Histological sections were obtained after dMRI. Optical microscopy photomicrographs of the optic nerves (stained with toluidine blue) are available, as well as their corresponding automatic segmentations of axons and myelin.

Measurement-oriented deep-learning workflow for improved segmentation of myelin and axons in high-resolution images of human cerebral white matter

BACKGROUND

Standard segmentation of high-contrast electron micrographs (EM) identifies myelin accurately but does not translate easily into measurements of individual axons and their myelin, even in cross-sections of parallel fibers. We describe automated segmentation and measurement of each myelinated axon and its sheath in EMs of arbitrarily oriented human white matter from autopsies.

NEW METHODS

Preliminary segmentation of myelin, axons and background by machine learning, using selected filters, precedes automated correction of systematic errors. Final segmentation is done by a deep neural network (DNN). Automated measurement of each putative fiber rejects measures encountering pre-defined artifacts and excludes fibers failing to satisfy pre-defined conditions.

RESULTS

Improved segmentation of three sets of 30 annotated images each (two sets from human prefrontal white matter and one from human optic nerve) is achieved with a DNN trained only with a subset of the first set from prefrontal white matter. Total number of myelinated axons identified by the DNN differed from expert segmentation by 0.2%, 2.9%, and -5.1%, respectively. G-ratios differed by 2.96%, 0.74% and 2.83%. Intraclass correlation coefficients between DNN and annotated segmentation were mostly >0.9, indicating nearly interchangeable performance.

COMPARISON WITH EXISTING METHOD(S)

Measurement-oriented studies of arbitrarily oriented fibers from central white matter are rare. Published methods are typically applied to cross-sections of fascicles and measure aggregated areas of myelin sheaths and axons, allowing estimation only of average g-ratio.

CONCLUSIONS

Automated segmentation and measurement of axons and myelin is complex. We report a feasible approach that has so far proven comparable to manual segmentation.

Histological validation of per-bundle water diffusion metrics within a region of fiber crossing following axonal degeneration

Micro-architectural characteristics of white matter can be inferred through analysis of diffusion-weighted magnetic resonance imaging (dMRI). The diffusion-dependent signal can be analyzed through several methods, with the tensor model being the most frequently used due to its straightforward interpretation and low requirements for acquisition parameters. While valuable information can be gained from the tensor-derived metrics in regions of homogeneous tissue organization, this model does not provide reliable microstructural information at crossing fiber regions, which are pervasive throughout human white matter. Several multiple fiber models have been proposed that seem to overcome the limitations of the tensor, with few providing per-bundle dMRI-derived metrics. However, biological interpretations of such metrics are limited by the lack of histological confirmation. To this end, we developed a straightforward biological validation framework. Unilateral retinal ischemia was induced in ten rats, which resulted in axonal (Wallerian) degeneration of the corresponding optic nerve, while the contralateral was left intact; the intact and injured axonal populations meet at the optic chiasm as they cross the midline, generating a fiber crossing region in which each population has different diffusion properties. Five rats served as controls. High-resolution ex vivo dMRI was acquired five weeks after experimental procedures. We correlated and compared histology to per-bundle descriptors derived from three methodologies for dMRI analysis (constrained spherical deconvolution and two multi-tensor representations). We found a tight correlation between axonal density (as evaluated through automatic segmentation of histological sections) with per-bundle apparent fiber density and fractional anisotropy (derived from dMRI). The multi-fiber methods explored were able to correctly identify the damaged fiber populations in a region of fiber crossings (chiasm). Our results provide validation of metrics that bring substantial and clinically useful information about white-matter tissue at crossing fiber regions. Our proposed framework is useful to validate other current and future dMRI methods.

Deep learning for high-throughput quantification of oligodendrocyte ensheathment at single-cell resolution

High-throughput quantification of oligodendrocyte myelination is a challenge that, if addressed, would facilitate the development of therapeutics to promote myelin protection and repair. Here, we established a high-throughput method to assess oligodendrocyte ensheathment in-vitro, combining nanofiber culture devices and automated imaging with a heuristic approach that informed the development of a deep learning analytic algorithm. The heuristic approach was developed by modeling general characteristics of oligodendrocyte ensheathments, while the deep learning neural network employed a UNet architecture and a single-cell training method to associate ensheathed segments with individual oligodendrocytes. Reliable extraction of multiple morphological parameters from individual cells, without heuristic approximations, allowed the UNet to match the accuracy of expert-human measurements. The capacity of this technology to perform multi-parametric analyses at the level of individual cells, while reducing manual labor and eliminating human variability, permits the detection of nuanced cellular differences to accelerate the discovery of new insights into oligodendrocyte physiology.

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.

Axons morphometry in the human spinal cord

Due to the technical challenges of large-scale microscopy and analysis, to date only limited knowledge has been made available about axon morphometry (diameter, shape, myelin thickness, volume fraction), thereby limiting our understanding of neuronal microstructure and slowing down research on neurodegenerative pathologies. This study addresses this knowledge gap by establishing a state-of-the-art acquisition and analysis framework for mapping axon morphometry, and providing the first comprehensive mapping of axon morphometry in the human spinal cord. We dissected, fixed and stained a human spinal cord with osmium tetroxide, and used a scanning electron microscope to image the entirety of 23 axial slices, covering C1 to L5 spinal levels. An automatic method based on deep learning was then used to segment each axon and myelin sheath to produce maps of axon morphometry. These maps were then registered to a standard spinal cord magnetic resonance imaging (MRI) template. Between 500,000 (lumbar) and 1 million (cervical) myelinated axons were segmented at each level of this human spinal cord. Morphometric features show a large disparity between tracts, but high right-left symmetry. Our results suggest a modality-based organization of the dorsal column in the human, as it has been observed in the rat. The generated axon morphometry template is publicly available at https://osf.io/8k7jr/ and could be used as a reference for quantitative MRI studies. The proposed framework for axon morphometry mapping could be extended to other parts of the central or peripheral nervous system that exhibit coherently-oriented axons.

Microstructural imaging in the spinal cord and validation strategies

In vivo histology using magnetic resonance imaging (MRI) is a newly emerging research field that aims to non-invasively characterize tissue microstructure. The implications of in vivo histology are many, from discovering novel biomarkers to studying human development, to providing tools for disease diagnosis and monitoring the effects of novel treatments on tissue. This review focuses on quantitative MRI (qMRI) techniques that are used to map spinal cord microstructure. Opening with a rationale for non-invasive imaging of the spinal cord, this article continues with a brief overview of the existing MRI techniques for axon and myelin imaging, followed by the specific challenges and potential solutions for acquiring and processing such data. The final part of this review focuses on histological validation, with suggested tissue preparation, acquisition and processing protocols for large-scale microscopy.

Myelin abnormality in Charcot-Marie-Tooth type 4J recapitulates features of acquired demyelination

OBJECTIVE

Charcot-Marie-Tooth type 4J (CMT4J) is a rare autosomal recessive neuropathy caused by mutations in FIG4 that result in loss of FIG4 protein. This study investigates the natural history and mechanisms of segmental demyelination in CMT4J.

METHODS

Over the past 9 years, we have enrolled and studied a cohort of 12 CMT4J patients, including 6 novel FIG4 mutations. We evaluated these patients and related mouse models using morphological, electrophysiological, and biochemical approaches.

RESULTS

We found sensory motor demyelinating polyneuropathy consistently in all patients. This underlying myelin pathology was associated with nonuniform slowing of conduction velocities, conduction block, and temporal dispersion on nerve conduction studies, which resemble those features in acquired demyelinating peripheral nerve diseases. Segmental demyelination was also confirmed in mice without Fig4 (Fig4-/- ). The demyelination was associated with an increase of Schwann cell dedifferentiation and macrophages in spinal roots where nerve-blood barriers are weak. Schwann cell dedifferentiation was induced by the increasing intracellular Ca2+ . Suppression of Ca2+ level by a chelator reduced dedifferentiation and demyelination of Schwann cells in vitro and in vivo. Interestingly, cell-specific knockout of Fig4 in mouse Schwann cells or neurons failed to cause segmental demyelination.

INTERPRETATION

Myelin change in CMT4J recapitulates the features of acquired demyelinating neuropathies. This pathology is not Schwann cell autonomous. Instead, it relates to systemic processes involving interactions of multiple cell types and abnormally elevated intracellular Ca2+ . Injection of a Ca2+ chelator into Fig4-/- mice improved segmental demyelination, thereby providing a therapeutic strategy against demyelination. Ann Neurol 2018;83:756-770.

AxonDeepSeg: automatic axon and myelin segmentation from microscopy data using convolutional neural networks

Segmentation of axon and myelin from microscopy images of the nervous system provides useful quantitative information about the tissue microstructure, such as axon density and myelin thickness. This could be used for instance to document cell morphometry across species, or to validate novel non-invasive quantitative magnetic resonance imaging techniques. Most currently-available segmentation algorithms are based on standard image processing and usually require multiple processing steps and/or parameter tuning by the user to adapt to different modalities. Moreover, only a few methods are publicly available. We introduce AxonDeepSeg, an open-source software that performs axon and myelin segmentation of microscopic images using deep learning. AxonDeepSeg features: (i) a convolutional neural network architecture; (ii) an easy training procedure to generate new models based on manually-labelled data and (iii) two ready-to-use models trained from scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Results show high pixel-wise accuracy across various species: 85% on rat SEM, 81% on human SEM, 95% on mice TEM and 84% on macaque TEM. Segmentation of a full rat spinal cord slice is computed and morphological metrics are extracted and compared against the literature. AxonDeepSeg is freely available at https://github.com/neuropoly/axondeepseg .

Axon and Myelin Morphology in Animal and Human Spinal Cord

Characterizing precisely the microstructure of axons, their density, size and myelination is of interest for the neuroscientific community, for example to help maximize the outcome of studies on white matter (WM) pathologies of the spinal cord (SC). The existence of a comprehensive and structured database of axonal measurements in healthy and disease models could help the validation of results obtained by different researchers. The purpose of this article is to provide such a database of healthy SC WM, to discuss the potential sources of variability and to suggest avenues for robust and accurate quantification of axon morphometry based on novel acquisition and processing techniques. The article is organized in three sections. The first section reviews morphometric results across species according to range of densities and counts of myelinated axons, axon diameter and myelin thickness, and characteristics of unmyelinated axons in different regions. The second section discusses the sources of variability across studies, such as age, sex, spinal pathways, spinal levels, statistical power and terminology in regard to tracts and protocols. The third section presents new techniques and perspectives that could benefit histology studies. For example, coherent anti-stokes Raman spectroscopy (CARS) imaging can provide sub-micrometric resolution without the need for fixation and staining, while slide scanners and stitching algorithms can provide full cross-sectional area of SC. In combination with these acquisition techniques, automatic segmentation algorithms for delineating axons and myelin sheath can help provide large-scale statistics on axon morphometry.

Is There Evidence for Myelin Modeling by Astrocytes in the Normal Adult Brain?

A set of astrocytic process associated with altered myelinated axons is described in the forebrain of normal adult rodents with confocal, electron microscopy, and 3D reconstructions. Each process consists of a protuberance that contains secretory organelles including numerous lysosomes which polarize and open next to disrupted myelinated axons. Because of the distinctive asymmetric organelle distribution and ubiquity throughout the forebrain neuropil, this enlargement is named paraxial process (PAP). The myelin envelope contiguous to the PAP displays focal disruption or disintegration. In routine electron microscopy clusters of large, confluent, lysosomes proved to be an effective landmark for PAP identification. In 3D assemblies lysosomes organize a series of interconnected saccules that open up to the plasmalemma next to the disrupted myelin envelope(s). Activity for acid hydrolases was visualized in lysosomes, and extracellularly at the PAP-myelin interface and/or between the glial and neuronal outer aspects. Organelles in astrocytic processes involved in digesting pyknotic cells and debris resemble those encountered in PAPs supporting a likewise lytic function of the later. Conversely, processes entangling tripartite synapses and glomeruli were devoid of lysosomes. Both oligodendrocytic and microglial processes were not associated with altered myelin envelopes. The possible roles of the PAP in myelin remodeling in the context of the oligodendrocyte-astrocyte interactions and in the astrocyte's secretory pathways are discussed.

Promise and pitfalls of g-ratio estimation with MRI

The fiber g-ratio is the ratio of the inner to the outer diameter of the myelin sheath of a myelinated axon. It has a limited dynamic range in healthy white matter, as it is optimized for speed of signal conduction, cellular energetics, and spatial constraints. In vivo imaging of the g-ratio in health and disease would greatly increase our knowledge of the nervous system and our ability to diagnose, monitor, and treat disease. MRI based g-ratio imaging was first conceived in 2011, and expanded to be feasible in full brain white matter with preliminary results in 2013. This manuscript reviews the growing g-ratio imaging literature and speculates on future applications. It details the methodology for imaging the g-ratio with MRI, and describes the known pitfalls and challenges in doing so.

Multiple uses of fibrin sealant for nervous system treatment following injury and disease

Lesions to the nervous system often produce hemorrhage and tissue loss that are difficult, if not impossible, to repair. Therefore, scar formation, inflammation and cavitation take place, expanding the lesion epicenter. This significantly worsens the patient conditions and impairment, increasing neuronal loss and glial reaction, which in turn further decreases the chances of a positive outcome. The possibility of using hemostatic substances that also function as a scaffold, such as the fibrin sealant, reduces surgical time and improve postoperative recovery. To date, several studies have demonstrated that human blood derived fibrin sealant produces positive effects in different interventions, becoming an efficient alternative to suturing. To provide an alternative to homologous fibrin sealants, the Center for the Study of Venoms and Venomous Animals (CEVAP, Brazil) has proposed a new bioproduct composed of certified animal components, including a thrombin-like enzyme obtained from snake venom and bubaline fibrinogen. Thus, the present review brings up to date literature assessment on the use of fibrin sealant for nervous system repair and positions the new heterologous bioproduct from CEVAP as an alternative to the commercial counterparts. In this way, clinical and pre-clinical data are discussed in different topics, ranging from central nervous system to peripheral nervous system applications, specifying positive results as well as future enhancements that are necessary for improving the use of fibrin sealant therapy.

Modeling white matter microstructure

Quantitative magnetic resonance imaging can be combined with advanced biophysical models to measure microstructural features of white matter. Non-invasive microstructural imaging has the potential to revolutionize neuroscience, and acquiring these measures in clinically feasible times would greatly improve patient monitoring and clinical studies of drug efficacy. However, a good understanding of microstructural imaging techniques is essential to set realistic expectations and to prevent over-interpretation of results. This review explains the methodology behind microstructural modeling and imaging, and gives an overview of the breakthroughs and challenges associated with it.

AxonPacking: An Open-Source Software to Simulate Arrangements of Axons in White Matter

HIGHLIGHTS

AxonPacking: Open-source software for simulating white matter microstructure.

Validation on a theoretical disk packing problem.

Reproducible and stable for various densities and diameter distributions.

Can be used to study interplay between myelin/fiber density and restricted fraction.

Quantitative Magnetic Resonance Imaging (MRI) can provide parameters that describe white matter microstructure, such as the fiber volume fraction (FVF), the myelin volume fraction (MVF) or the axon volume fraction (AVF) via the fraction of restricted water (fr). While already being used for clinical application, the complex interplay between these parameters requires thorough validation via simulations. These simulations required a realistic, controlled and adaptable model of the white matter axons with the surrounding myelin sheath. While there already exist useful algorithms to perform this task, none of them combine optimisation of axon packing, presence of myelin sheath and availability as free and open source software. Here, we introduce a novel disk packing algorithm that addresses these issues. The performance of the algorithm is tested in term of reproducibility over 50 runs, resulting density, and stability over iterations. This tool was then used to derive multiple values of FVF and to study the impact of this parameter on fr and MVF in light of the known microstructure based on histology sample. The standard deviation of the axon density over runs was lower than 10⁻³ and the expected hexagonal packing for monodisperse disks was obtained with a density close to the optimal density (obtained: 0.892, theoretical: 0.907). Using an FVF ranging within [0.58, 0.82] and a mean inter-axon gap ranging within [0.1, 1.1] μm, MVF ranged within [0.32, 0.44] and fr ranged within [0.39, 0.71], which is consistent with the histology. The proposed algorithm is implemented in the open-source software AxonPacking (https://github.com/neuropoly/axonpacking) and can be useful for validating diffusion models as well as for enabling researchers to study the interplay between microstructure parameters when evaluating qMRI methods.