Axons, the long cable-like projections that carry signals between neurons, are fundamental to the structural integrity and function of the nervous system. Measuring their density and organization is central to understanding diseases such as multiple sclerosis, in which axonal loss drives progressive disability, as well as Alzheimer’s disease and traumatic brain injury. The Bielschowsky silver stain, first introduced in 1902, remains a workhorse of clinical neuropathology because it renders thin axons and neurofibrils dark against a paler background. But converting these images into reliable numbers has long been a bottleneck. Manual counting and stereological sampling are slow and scale poorly, while shortcuts such as optical density measurements are confounded by stain quality and by silver deposits in non-axonal tissue. Line-intersection counting, another common surrogate, depends on arbitrary line placement and local fiber orientation.

Classical image-processing approaches fare no better. Color-deconvolution techniques, which work reasonably well on brightly colored immunohistochemical stains, perform poorly on the monochromatic, heterogeneous contrast of silver impregnations. Existing deep learning pipelines for axon segmentation, such as AxonDeepSeg and DeepACSON, were built for electron microscopy, immunofluorescence, or myelin-specific stains, and none had been optimized for brightfield silver-stained histology. BasNet was designed specifically to fill that gap, trained entirely on manually annotated Bielschowsky-stained tissue.

The architecture is an attention U-Net, an encoder-decoder convolutional network in which so-called attention gates selectively emphasize task-relevant features while suppressing background noise. The encoder compresses the image through four hierarchical levels, expanding from 32 to 256 channels with each downsampling step, while the decoder reconstructs a full-resolution segmentation map. At every skip connection, an attention gate computes spatial attention coefficients that weight encoder features according to their relevance, a mechanism that proved particularly valuable in silver-stained tissue, where non-specific background silver deposition can trigger false positives in a standard U-Net. The complete network contains approximately 7.85 million trainable parameters and processes 256 by 256 pixel RGB patches, outputting a binary mask in which axon pixels are distinguished from background.

Training presented two classic challenges: extreme class imbalance, with axons occupying only about 30 percent of the pixels, and the need to recover very thin, elongated structures. The team addressed both with Focal Tversky Loss, a function that penalizes false negatives more heavily than false positives and applies a quadratic focusing factor that concentrates learning on difficult boundary regions. This asymmetry was deliberate. For axon density quantification, a missed axon directly reduces the density estimate, whereas slight over-segmentation is less harmful. The design shows in the results: mean recall of 0.754 consistently exceeded mean precision of 0.717 on the test set. An ablation study confirmed that the attention gates improved mean test Dice from 0.699 to 0.715, a small but consistent gain achieved at a cost of only about one percent additional parameters.

The training data were laborious to produce. Two medical doctors annotated 33 image tiles covering more than 25 million pixels, drawn from 26 whole-slide images spanning gray matter, white matter, demyelinated lesions, and background regions from the brains of four multiple sclerosis patients obtained through the German MS Brain Bank. Slides were prepared by different technicians at different times, deliberately capturing the real-world staining variability that plagues clinical material. The tiles were subdivided into more than 86,000 augmented training patches, and a leave-one-WSI-out cross-validation across twelve folds confirmed that no single training slide was disproportionately critical to performance, with Dice scores ranging narrowly between 0.687 and 0.725.

The most telling evaluation was a direct comparison with human variability. When two independent raters segmented the same two tiles, their mutual agreement, measured by the F1 or Dice coefficient, was 0.637 and 0.671. The model’s agreement with each individual rater ranged from 0.681 to 0.736, meeting or exceeding the human-to-human ceiling on both tiles. Comparable benchmarks in the literature, such as inter-rater Dice scores of around 0.78 for optic nerve axons and an average Jaccard index of 0.653 for nuclei segmentation, place BasNet squarely within the range of inter-observer disagreement. The authors conclude that the principal remaining source of error is annotation ambiguity rather than model limitation. On the held-out test set the model achieved a mean Dice of 0.717 and an intersection over union of 0.584, though performance varied widely across tiles, from perfect scores on a background tile to a low of 0.392 on a poorly stained slide, mirroring the difficulties that poor staining quality also poses to human annotators.

Beyond tile-level scores, the team demonstrated a pipeline that scales to whole slides. Aggregated segmentation masks produce spatially resolved maps of axon density at 70 micrometer resolution, revealing the expected anatomy: uniformly bright white matter tracts, a dimmer and more variable cortical ribbon, and visibly reduced density in multiple sclerosis lesions. A structural tensor analysis, built on image gradients and eigen decomposition, additionally yields maps of dominant fiber orientation and local coherency, a fractional anisotropy-like measure of directional alignment. Highly organized fiber bundles appear bright and smoothly colored, while crossing-fiber zones, cortex, and lesion borders show low coherency, consistent with known disruption of the tissue’s myeloarchitecture. These outputs provide quantitative readouts suitable for downstream statistical analysis and cohort comparison, in place of arbitrary sampling decisions.

As a benchmark comparison, the team trained nnU-Net v2, a widely used self-configuring segmentation framework, on the identical data. Their compact attention U-Net achieved a higher mean Dice on both the full eight-tile test set, 0.717 versus 0.555, and the seven axon-bearing tiles, 0.677 versus 0.634, while using roughly six times fewer parameters than nnU-Net’s self-configured network. The advantage was clearest on a white matter tile where axons run perpendicular to the cutting plane and appear as small cross-sectional profiles. The authors caution that with only eight test tiles, this comparison should be read as an honest benchmark rather than a definitive ranking.

The study has clear limitations. All tissue came from a single disease context, multiple sclerosis, and followed a single staining protocol, and all training data were digitized at one resolution of 0.273 micrometers per pixel. Preliminary tests at coarser resolutions looked visually convincing but were not quantitatively evaluated, and generalization to other silver impregnation methods or other laboratories remains unverified. The annotation set of 33 tiles is small, and the inter-rater ceiling was estimated from just two tiles. Nevertheless, the team has released the full pipeline as open-source software, with the trained model hosted on Hugging Face and the source code archived on Zenodo, making it immediately deployable on tiled images and whole-slide scans. The researchers envision future integration with multimodal registration frameworks that would align these histological orientation maps with ex vivo and in vivo MRI data, establishing a quantitative histological ground truth for microstructure imaging and opening a new bridge between the microscopic world of silver-stained axons and the macroscopic images used in clinical neurology.

Subject of Research: Automated deep learning segmentation of axons in Bielschowsky silver-stained histological brain sections

Article Title: BasNet: Attention U-Net-Based Automated Axon Segmentation in Bielschowsky Silver-Stained Histology

Article References: Schönenberger, L., Egli, L., Gkotsoulias, D., Stadelmann, C., & Granziera, C. (2026). BasNet: Attention U-Net-Based Automated Axon Segmentation in Bielschowsky Silver-Stained Histology. Neuroinformatics, 24(3), Article 59. https://doi.org/10.1007/s12021-026-09815-z

Image Credits: AI Generated

DOI: 10.1007/s12021-026-09815-z

Keywords: Bielschowsky silver stain, axon segmentation, attention U-Net, deep learning, computational pathology, multiple sclerosis, neuropathology, image segmentation, axon density, whole-slide imaging, Focal Tversky Loss, Neuroinformatics

The study addresses a pressing challenge in the medical community: the early detection and classification of Alzheimer’s Disease and its precursors. Traditional diagnostic methods often rely on subjective assessments and can be influenced by various factors, leading to potential misdiagnoses or delayed treatments. The authors’ innovative approach employs deep learning algorithms, harnessing the power of artificial intelligence (AI) to enhance diagnostic accuracy, thereby revolutionizing the landscape of neurodegenerative disease diagnostics.

By integrating deep learning with CT imaging, the researchers developed a model capable of analyzing intricate patterns within brain scans that may not be immediately visible to human eyes. This AI-driven model was trained on a dataset comprising thousands of CT images, allowing it to learn the subtle differences indicative of HC, MCI, and AD. The methodology also tackled the inherent variability in human brain anatomy and the stage of disease which can complicate the diagnostic process. The researchers’ model promises to provide a robust solution to these complexities.

An impressive aspect of the study is its emphasis on the explainability of the AI model. The researchers prioritized not only accuracy but also the interpretability of the findings. Understanding the reasons behind a model’s predictions can significantly aid clinicians in making more informed decisions regarding patient care. By employing techniques such as heatmaps, the researchers could visually represent the specific areas of the brain that contributed most significantly to the classification, thus providing essential insights for clinical practitioners.

Furthermore, the research highlights the efficiency of deep learning algorithms in processing large datasets. Given the rising prevalence of neurodegenerative diseases globally, the need for scalable and cost-effective diagnostic tools has never been more critical. With the capacity to analyze thousands of images within a fraction of the time it would take a human, this study underscores the transformative potential of AI in medicine.

Additionally, the implications of this research extend beyond mere classification. With the enhancement of diagnostic capabilities, there is a corresponding hope for improving patient outcomes through earlier detection and customized treatment strategies. The study can pave the way for proactive monitoring of individuals at risk for cognitive decline, enabling timely interventions that may slow disease progression and enhance the quality of life.

As the world grapples with an aging population and the accompanying rise in age-related illnesses, such advanced methodologies in medical diagnostics are essential. The adoption of AI tools such as those developed in this study could signify a paradigm shift in how neurological conditions are diagnosed and treated. It lays the groundwork for future research, pushing the boundaries of current understanding and fostering a more personalized approach to patient care.

Moreover, as healthcare systems around the globe continue to evolve, the integration of AI in clinical workflows highlights the importance of collaboration between technologists and healthcare professionals. Such partnerships are crucial to ensure that the tools developed are not only scientifically sound but also applicable in real-world settings. The researchers advocate for continuous collaboration to refine these models and validate their applicability across diverse populations.

Despite the promising results exhibited in this study, the authors emphasize the necessity of continuous improvement and verification of the technology. They advocate for larger-scale studies that encompass varied demographics to further explore the efficacy of the AI model in different populations. This step is crucial for ensuring that the technology is both widely applicable and sensitive to the biological diversity observed in human populations.

In conclusion, the study by Hsiao and colleagues represents a pivotal moment in the intersection of artificial intelligence and medical diagnostics. It highlights how technology can be leveraged to better understand complex medical conditions and fosters hope for more effective interventions. As we move further into the era of personalized medicine, the ability to adopt cutting-edge technology into clinical practices will be paramount.

In a world where the intersection of technology and health is becoming increasingly intertwined, the advancements made in this study could very well herald a new age of diagnostic precision. Organizing future efforts towards refining these tools will surely remain critical in the years to come. The ongoing research and discussions around such innovations will continue to energize the scientific community and inspire new pathways for treating neurodegenerative diseases.

As neuroimaging techniques and AI continue to evolve, we can anticipate even more groundbreaking studies that could further enrich our understanding of the human brain and the complexities of cognitive impairments. The potential for these technologies to influence not just diagnosis, but the broader landscape of healthcare, is immense.

The ongoing collaboration between scientists and clinicians will be instrumental in closing the gap between research and practical application. As the capabilities of deep learning algorithms expand, the future of diagnosing and managing neurodegenerative conditions seems brighter than ever, allowing for a proactive rather than reactive approach to cognitive health. The dissemination of such research is vital, as it encourages a communal effort towards understanding, diagnosing, and ultimately treating cognitive disorders that affect millions worldwide.

Subject of Research: Classification of Healthy Control, Mild Cognitive Impairment, and Alzheimer’s Disease using Deep Learning and CT Imaging.

Article Title: Classification of HC, MCI, and AD Based on CT Using Deep Learning.

Article References:

Hsiao, IT., Lin, KJ., Chang, CC. et al. Classification of HC, MCI, and AD Based on CT Using Deep Learning.
J. Med. Biol. Eng. (2025). https://doi.org/10.1007/s40846-025-00985-w

Image Credits: AI Generated

DOI:

Keywords: Neuroimaging, Deep Learning, Alzheimer’s Disease, Mild Cognitive Impairment, AI in Healthcare, Medical Diagnostics, CT Imaging, Brain Health.