Round up: AI co-pilot boosts non-invasive brain-computer interface, and more
Neuro Rehab Times explores the latest developments in the world of neurology and neurorehabilitation.
Gut immune system mouse model of Alzheimer’s provided new target for therapeutics
The gut immune system is altered in a mouse model of Alzheimer’s, new research has found.
The research shows that some immune cells in the gut travel along the brain/gut axis in a mouse model of Alzheimer’s disease (AD), a finding that provides a potential new therapeutic pathway for the memory-robbing malady.
It also reveals that feeding the mice a high fiber diet reduces AD-related frailty, including tremor.
The team found that specific antibody-producing B cells, normally responsible for keeping the microbiome and the gut immune system in harmony, were reduced in the mice bred to develop AD.
It was also discovered that this cell type has a migratory signature; researchers found the gut-specific B cells and their migratory receptors in the brain and in its border region, the meningeal dura mater.
“Remarkably, we found that these immune cells in the brain border which recognise bacteria living in the intestines were accumulating in the AD brain,” said Priya Makhijani, postdoctoral fellow and immunologist who led the research at the Buck Institute.
The team found that this gut immune cell receptor’s binding partner, a well-studied chemokine known for migration, was produced at higher levels in the glia, the inflammatory cells in the AD brain.
The migratory signature was also identified in human AD brains via data mining of previously conducted studies. Working with collaborators at the University Health Network, part of the University of Toronto, the team conducted blocking experiments in the axis using a small molecule drug, suggesting that a new long-range mechanism might be acting along the gut-brain axis.
The team also found that feeding the animals the anti-inflammatory pre-biotic fiber inulin restored balance in the gut of the AD mice.
“We found these migrating cells were replenished in the gut and that AD-related frailty, including the tremor trait, was reduced in the animals,” said Makhijani, noting that insulin makes short chain fatty acids and other metabolites that concentrate in the gut and can also circulate systemically.
Makhijani says the diet improved gut health and reduced chemokine signalling in the brain.
The team say that while the high fiber diet did not consistently reduce the levels of plaques in the mice’s brain, it did impact overall wellbeing.
While the study provides a comprehensive characterization of gut immune system changes in a neurological disease, researchers say more work is needed to see if those changes are a response to brain alterations or whether they drive the disease itself.
Speech-based model detects early neurological disorders
Researchers have developed a novel deep learning framework that significantly improves the accuracy and interpretability of detecting neurological disorders through speech.
Automated speech analysis offers high efficiency, low cost, and non-invasiveness. However, current mainstream methods often suffer from over-reliance on handcrafted features, limited capacity to model temporal-variable interactions, and poor interpretability.
To address these challenges, the research team the Institute of Health and Medical Technology, the Hefei Institutes of Physical Science of the Chinese Academy of Sciences, proposed Cross-Time and Cross-Axis Interactive Transformer (CTCAIT) for multivariate time series analysis.
This framework first employs a large-scale audio model to extract high-dimensional temporal features from speech, representing them as multidimensional embeddings along time and feature axes. It then leverages the Inception Time network to capture multi-scale and multi-level patterns within the time series.
By integrating cross-time and cross-channel multi-head attention mechanisms, CTCAIT effectively captures pathological speech signatures embedded across different dimensions.
The method achieved a detection accuracy of 92.06 per cent on a Mandarin Chinese dataset and 87.73 per cent on an external English dataset, demonstrating strong cross-linguistic generalisability.
Brain signaling that sets Parkinson’s disease apart from essential tremor revealed
Researchers have identified a neurochemical signature that sets Parkinson’s disease apart from essential tremor – two of the most common movement disorders, but each linked to distinct changes in the brain.
The team identified unique chemical signaling patterns of two key neurotransmitters — dopamine and serotonin — that distinguish these two disorders.
“This study builds on decades of work,” said Read Montague, a scientist at the Fralin Biomedical Research Institute and a co-senior author, who with colleagues developed the multi-faceted technologies and the theoretical constructs for the work over their 15 years at the research institute.
The researchers focused on a brain region involved in decision-making and reward processing – the caudate of the striatum.
Using a machine learning-enhanced electrochemical technique during deep brain stimulation (DBS) surgery on essential tremor and Parkinson’s patients, the team measured fast fluctuations as patients played a game involving fair and unfair offers, a task designed to understand decision-making and brain chemistry.
In an early study to emerge from this work in 2018, researchers revealed the first-ever recordings of simultaneous sub-second fluctuations of dopamine and serotonin during active decision-making in a conscious human subject.
During DBS surgeries in 2017 and 2018, Wake Forest University neurosurgeons Adrian Laxton and Stephen Tatter helped perform the recordings using carbon fiber electrodes in Parkinson’s disease and essential tremor patients while they played a game where they accepted or rejected offers.
The research protocol was carried during a part of the surgery where the neurosurgeons already monitor brain activity in real time to precisely locate a small target area of the brain for electrical stimulation to treat the symptoms of the disorders.
Now, in the new study, researchers applied a computational model to track how those patients formed and adjusted their expectations during the game, revealing signature chemical signaling patterns tied to each disorder.
In people with essential tremor, monetary offers that violated their expectations during the game triggered a seesaw pattern: dopamine levels rose, while serotonin dropped.
This oppositional response – with one neurotransmitter rising as the other fell – mirrored patterns seen in earlier studies of brain activity during decision-making.
In the latest findings, this reciprocal neurochemical signaling was absent in patients with Parkinson’s disease.
It’s known that dopamine-producing neurons die in Parkinson’s disease, so researchers expected dopamine to be the clearest chemical difference in the brain.
But when they looked closely using refined tools and a model of how people formed expectations, it was not dopamine that best distinguished Parkinson’s from essential tremor.
Instead, it was serotonin, a different neurotransmitter that has not been as prominent in theories of Parkinson’s disease, opening a new view and potentially powerful scientific and clinical insight into this disease.
The computational model followed a form of machine learning known as reinforcement learning, which gradually improved its ability to detect patterns as it processed more data from past experiments.
The team did experiments in mice to inform the approach, which helped Sands refine and apply the statistical model to extract new insights from human patient decision-making behaviour.
Researchers saw that certain prediction errors – mismatches between what the research subjects expected and what they received – evoked changes in serotonin activity that were strong indicators of which disease the patient had.
“It’s very powerful to link moment-to-moment changes in internal beliefs — here what a person expects from others — to measurable chemical signals in the brain,” said Dan Bang, an associate professor at the Center of Functionally Integrative Neuroscience at Aarhus University in Denmark, adjunct associate professor at the Fralin Biomedical Research Institute, and one of the study authors.
“This opens a new window into how deeply human cognitive processes, like social evaluation, are shaped by disease.”
While the findings offer new insight into how Parkinson’s disease and essential tremor differ at the chemical level, the researchers see this as just the beginning.
AI co-pilot boosts non-invasive brain-computer interface
Engineers have developed a wearable, non-invasive brain-computer interface system that utilises AI as a co-pilot to help infer user intent and complete tasks by moving a robotic arm or a computer cursor.
The study shows that the interface demonstrates a new level of performance in non-invasive brain-computer interface, or BCI, systems.
This could lead to a range of technologies to help people with limited physical capabilities, such as those with paralysis or neurological conditions, handle and move objects more easily and precisely.
The team at UCLA developed custom algorithms to decode electroencephalography, or EEG – a method of recording the brain’s electrical activity – and extract signals that reflect movement intentions.
They paired the decoded signals with a camera-based artificial intelligence platform that interprets user direction and intent in real time. The system allows individuals to complete tasks significantly faster than without AI assistance.
“By using artificial intelligence to complement brain-computer interface systems, we’re aiming for much less risky and invasive avenues,” said study leader Jonathan Kao, an associate professor of electrical and computer engineering at the UCLA Samueli School of Engineering.
“Ultimately, we want to develop AI-BCI systems that offer shared autonomy, allowing people with movement disorders, such as paralysis or ALS, to regain some independence for everyday tasks.”
State-of-the-art, surgically implanted BCI devices can translate brain signals into commands, but the benefits they currently offer are outweighed by the risks and costs associated with neurosurgery to implant them.
More than two decades after they were first demonstrated, such devices are still limited to small pilot clinical trials. Meanwhile, wearable and other external BCIs have demonstrated a lower level of performance in detecting brain signals reliably.
To address these limitations, the researchers tested their new non-invasive AI-assisted BCI with four participants – three without motor impairments and a fourth who was paralysed from the waist down.
The BCI deciphered electrical brain signals that encoded the participants’ intended actions.
Using a computer vision system, the custom-built AI inferred the users’ intent – not their eye movements – to guide the cursor and position the blocks.
“Next steps for AI-BCI systems could include the development of more advanced co-pilots that move robotic arms with more speed and precision, and offer a deft touch that adapts to the object the user wants to grasp,” said co-lead author Johannes Lee, a UCLA electrical and computer engineering doctoral candidate advised by Kao.
“And adding in larger-scale training data could also help the AI collaborate on more complex tasks, as well as improve EEG decoding itself.”
FDA approves Leqembi subcutaneous injection for early Alzheimer’s disease
The US Food and Drug Administration (FDA) has approved the Biologics License Application (BLA) for once weekly lecanemab subcutaneous injection for maintenance dosing for the treatment of early Alzheimer’s disease.
The US brand name for the subcutaneous autoinjector is LEQEMBI IQLIK (pronounced “I Click”).
“Eisai’s continued work to support and simplify patient and healthcare administration and treatment is an important work to help remove potential bottlenecks in healthcare and broaden patient population while supporting a sustainable long-term cost of treatment,” says Gunilla Osswald, CEO at BioArctic.
