New Microscopy and AI Bring Hope for Early Detection of Neurological Diseases

Seeing the Invisible

Recent advances in imaging allow researchers to view brain cells with unprecedented clarity. This capability opens a new window into neurological disease. At the Presidential Symposium on October 18, during the 151st Annual Meeting of the American Academy of Neurology, experts presented how advanced microscopes and AI are changing disease identification.

Dimitri Krainc, M. D., Ph. D., FANA, president of the Academy and chair of Neurology at Northwestern University in Chicago, emphasized that neurological diseases stem from changes in cells and their complex structures. Until recently, many of these transformations were invisible to standard methods. New technologies provide a dynamic, detailed view of disease biology. They can reveal anomalies earlier and guide treatments toward underlying molecular disruptions.

Super-Resolution Microscopy

Craig Blackstone, M. D., Ph. D., FANA, a professor at Harvard Medical School, presented how super-resolution microscopy changes the questions we can ask about neurological diseases. He examines the endoplasmic reticulum, a network of organelles essential for cell function. These techniques show that parts once thought to be flat sheets are actually dynamic networks of small tubes. This discovery offers a new perspective on how cells function and what goes wrong in protein mutation-related disorders, including hereditary spastic paraplegias. When we can see how structures inside neurons are organized and how they change dynamically in disease, we can link a genetic or molecular anomaly to what happens inside the cell. This uncovers previously invisible mechanisms and identifies new intervention points.

Live imaging allows researchers to track organelles in real time. They can observe how mitochondria produce and distribute energy. They also see how cells remove damaged mitochondria through autophagy, a process like clearing out an old sofa that is too big for the regular trash bag. Erika Holzbaur, Ph. D., a professor at the University of Pennsylvania, described these images as „like magic, like a timelapse video of a city at night.”These live films provide essential clues to understand basic mechanisms and how they are altered in disease.

Live Imaging of Organelles

Holzbaur's team used live imaging to discover an unexpected link between Fragile X syndrome and mitochondria. Studies on mouse and human neurons showed that in the absence of the FMRP protein, mitochondria function poorly. This opens new pathways for treatments. Holzbaur emphasized that when a neuron dies in a patient with ALS, Parkinson's, or Huntington's, it is too late to save it. The goal is to identify early signs of cellular stress to understand why and to save neurons before degeneration.

Diagnosing ALS remains difficult due to the lack of a definitive test. The process relies on a neurologist recognizing a pattern of symptoms and ruling out other conditions. Patients may spend nearly a year seeing doctors before confirmation. This means losing precious time to start available treatments or enroll in clinical trials, which are most useful in early stages.

Robotic microscopy combined with artificial intelligence can offer a faster, more objective diagnostic path. These thinkingmicroscopes can track thousands of motor neurons generated in the lab from patient skin or blood cells. They record appearance, function, and changes over time. Using AI, the massive image collection can be analyzed to detect disease-related patterns that are hard to identify manually. It also helps design experiments and learn from results.

Accelerating Diagnosis

Steven Finkbeiner, M. D., Ph. D., presented cutting-edge technologies for generating and analyzing imaging and genetic data. He noted that an AI model, trained exclusively on the appearance and behavior of patient-derived motor neurons, could reliably distinguish ALS neurons from healthy ones. This opens the possibility of faster, objective diagnosis when the clinical picture is unclear, instead of waiting months for symptoms to fully manifest.

AI also classified ALS neurons into significant subtypes. It tracked one subtype to a problem with protein movement in and out of the cell nucleus, providing clues about what drives the disease. In separate research, Finkbeiner's team used AI to automatically identify and measure signs of Alzheimer's disease in post-mortem brain tissue. These signs correlate with dementia severity, allowing rapid analysis of large brain sample collections. This approach is not specific to one disease; the same imaging and AI tools can be widely used for ALS, Alzheimer's, Parkinson's, Huntington's, and frontotemporal dementia. Although the work is in early stages, a technology developed to answer a question in one disease often accelerates discoveries in others.