New generations of high-resolution microscopy are giving researchers increasingly detailed views of what happens inside neurons during neurological diseases such as ALS, Parkinson’s and Alzheimer’s. Scientists can now observe structures inside living cells in real time and combine large volumes of microscopy images with artificial intelligence. The aim is to detect cellular abnormalities before neurons become irreversibly damaged or die.
Researchers presented several of these developments at the 151st Annual Meeting of the American Neurological Association (ANA 2026). Although the technologies remain primarily research tools, they could improve understanding of disease mechanisms and eventually contribute to earlier diagnosis and more targeted treatments. “Neurological diseases ultimately arise from changes in cells and the extraordinarily complex structures within them,” said ANA president Dimitri Krainc of Northwestern University Feinberg School of Medicine. According to Krainc, many of these changes could not be visualised until recently.
Looking deeper inside neurons
One approach is super-resolution microscopy, which enables scientists to examine structures beyond the resolution limits of conventional light microscopy. Craig Blackstone of Harvard Medical School and the Mass General Brigham Neuroscience Institute is using these techniques to study the endoplasmic reticulum (ER), a network of membranes inside cells that performs several essential functions. Advanced microscopy has revealed that structures previously regarded as relatively static sheets can actually form dynamic networks of extremely small tubes. Understanding how this architecture changes could be important in inherited neurological disorders such as hereditary spastic paraplegias, some of which are associated with mutations in proteins involved in shaping the ER.
This level of detail is particularly relevant for neurons. These highly specialised cells can extend over long distances and must maintain their internal organisation throughout a person’s lifetime. Live-cell microscopy adds another dimension by showing how structures behave over time. Researchers can follow mitochondria as they provide energy to different parts of a neuron and observe autophagy, the cellular process used to remove damaged components.
Erika Holzbaur of the University of Pennsylvania has used live imaging to investigate mitochondrial function and neuronal health. Her team found a link between fragile X syndrome and mitochondrial dysfunction in studies involving mouse and human neurons. Without FMRP, the protein missing in people with fragile X syndrome, mitochondria did not function normally. The broader objective is to identify cellular stress before neurons degenerate. By the time neurons die in diseases such as ALS, Parkinson’s or Huntington’s, opportunities to preserve those cells may already have been lost.
AI learns to recognise ALS cells
Combining microscopy with AI could also reveal disease patterns that are difficult for researchers to recognise manually. Steven Finkbeiner and colleagues are using robotic microscopy to monitor thousands of motor neurons generated in the laboratory from patients’ blood or skin cells. The system repeatedly records how individual cells look, function and change. AI algorithms can subsequently analyse the large image datasets. According to Finkbeiner, an AI model trained solely on the appearance and behaviour of patient-derived motor neurons was able to distinguish ALS neurons from healthy cells.
The model also separated ALS neurons into different subtypes. One subtype was associated with abnormalities in the transport of proteins into and out of the cell nucleus, potentially providing information about biological mechanisms underlying the disease. This is particularly relevant because ALS currently has no single definitive diagnostic test. Diagnosis is based on clinical assessment, recognising patterns of symptoms and excluding other possible conditions. The process can therefore take months.
Robotic microscopy combined with AI could eventually provide an additional, more objective source of information when a diagnosis is uncertain. However, the work remains at an early research stage. The ability to distinguish laboratory-grown neurons does not yet mean that AI microscopy can diagnose ALS in clinical practice. The same research group has applied AI to postmortem brain tissue from people with Alzheimer’s disease. Algorithms automatically identified and quantified pathological features that correlated with dementia severity, potentially enabling researchers to analyse large collections of tissue samples more efficiently.
Neural connections reconstructed in 3D
Other microscopy innovations are expanding the amount of structural information researchers can obtain from brain tissue. We previously reported on a high-resolution microscopy technique developed by researchers at the University of California, Berkeley. The method, called LICONN, makes it possible to visualise neural connections at nanoscale resolution while retaining molecular information about individual cells. The researchers demonstrated the technique by analysing brain tissue from mice. Rather than relying exclusively on electron microscopy to reconstruct extremely small neuronal structures, LICONN uses advanced light microscopy combined with expansion microscopy. Tissue is physically enlarged so that structures that would normally be too close together can be distinguished more clearly.
Researchers reconstructed approximately 1.5 million synapses and were able to map neural connections in three dimensions. Because molecular labels can be retained, the technique can provide information not only about where connections are located, but also about the biological identity of cells and proteins involved. The technology could help scientists investigate how changes in neural networks relate to brain function and disease. As with the microscopy approaches presented at ANA 2026, its current importance lies primarily in research rather than clinical diagnosis.
Earlier intervention
Together, these developments show us how microscopy is moving beyond producing increasingly detailed static images. Researchers can now reconstruct neural networks in 3D, watch organelles move inside living neurons and use AI to search millions of image features for patterns associated with disease. The potential clinical value lies in timing. Many neurological diseases are diagnosed after substantial cellular damage has already occurred. If microscopy can reveal molecular or structural abnormalities at an earlier stage, researchers may be able to identify new biomarkers and therapeutic targets before neurons are irreversibly lost.
There is still a substantial gap between observing early disease processes in laboratory cells or tissue and detecting them reliably in patients. Clinical validation will therefore be essential. For now, the rapidly improving resolution, automation and analytical capabilities of microscopy are providing researchers with a much more detailed view of where neurological disease begins.
References
Add ICT&health on Google
Show more content from ICT&health in Google Search.
