The Neurotech Breakthroughs That Are Reversing Memory Loss in Alzheimer's Patients
From flickering lights to electrode-studded brain implants, the tools fighting Alzheimer's are no longer just pharmaceuticals — and some of them are working.
For decades, Alzheimer’s research felt like a long, slow defeat. The disease would take a person’s memories, then their personality, then their independence — and the best medicine could do was hand out drugs that delayed the inevitable by a few months. That era isn’t over, but something has genuinely shifted. A combination of new pharmacology and, more interesting to us here at NeurotechMag, a surge of neurotechnology-driven approaches are producing results that would have sounded like science fiction ten years ago. Mice navigating mazes they couldn’t previously solve. Hippocampal atrophy halting in its tracks. Memory circuitry being switched back on. 🧠
None of this is a cure yet — let’s be clear about that from the start. But the pipeline is fuller, more diverse, and more mechanistically credible than it has ever been. And the shift in thinking, from “slow the decline” to “restore function,” is real. Here’s what’s actually happening.
The drug front: clearing the amyloid mess
The most talked-about development in Alzheimer’s pharmacology over the last two years is the arrival of anti-amyloid immunotherapy as a real, FDA-approved treatment category. Lecanemab (brand name Leqembi) and donanemab (Kisunla) received FDA approval in 2023 and 2024 respectively, and both have been shown to clear the majority of beta-amyloid plaque deposits from the brain over 12 to 18 months of IV infusion treatment, slowing cognitive decline by roughly 30%. That sounds modest. It isn’t. For early-stage patients, 30% slower decline is the difference between recognizing your grandchildren for an extra year or not.
In the pivotal TRAILBLAZER-ALZ 2 trial, donanemab slowed cognitive decline by 35% in early-stage patients, with some subgroups showing a 39% lower risk of disease progression. Eli Lilly also made donanemab the first amyloid therapy where treatment discontinuation is supported by evidence once the plaques are cleared — a meaningful practical win, since it potentially reduces both cost and the burden of repeated hospital infusions.
The next generation isn’t standing still either. Key developments include:
Trontinemab, which uses a “Brainshuttle” technology to push higher concentrations of antibody across the blood-brain barrier than previous drugs manage
Blarcamesine (Anavex’s oral candidate), which activates the sigma-1 receptor to drive neurons to clear plaques themselves — currently in Phase 2/3 trials
ALZ-801, targeting the upstream formation of toxic amyloid rather than the plaques themselves
NU-9, a small molecule from Northwestern University that clears toxic amyloid beta oligomers in hippocampal brain cells, the region most critical for learning and memory 🔬
UC San Francisco’s Adam Boxer compares where Alzheimer’s medication is today to the state of HIV drugs in the 1980s — early, imperfect, and full of side effects, but the beginning of a trajectory that eventually produced highly effective treatments. That framing is exactly right, and it’s a useful antidote to both overclaiming and despair.
Does the current approval of lecanemab and donanemab excite you, or do you think the modest effect size means the field is still fundamentally stuck? Drop your take in the comments.
The gamma wave gambit
Here’s where neurotechnology gets genuinely strange and wonderful. 💡 The MIT Picower Institute has spent years investigating a striking observation: Alzheimer’s patients show a measurable deficit in 40 Hz gamma oscillations, the fast rhythmic brain activity associated with memory encoding and retrieval. The question was whether you could just... put those oscillations back in, non-invasively, using lights and sound.
Turns out, you might be able to. A non-invasive therapy called GENUS (Gamma Entrainment Using Sensory stimulation), which uses 40 Hz audiovisual stimulation, has shown memory benefits in both Alzheimer’s mouse models and patients with mild probable Alzheimer’s disease. The proposed mechanisms are multiple:
Reduced amyloid and tau load, primarily by activating microglia (the brain’s immune cleanup crew)
Enhanced brain drainage and clearance of waste proteins
Restoration of neural network synchronization in memory-critical regions
Improved long-term potentiation — essentially, better capacity for the brain to form new memories ⚡
An open-label extension study published in Alzheimer’s & Dementia in 2025 evaluated the long-term effects of daily 40 Hz audiovisual stimulation on cognition and biomarkers in patients with mild Alzheimer’s disease. The MIT group is now running a prevention trial at Massachusetts General Hospital, recruiting adults aged 55+ with a family history of Alzheimer’s, using the device for 60 minutes a day at home for 12 months.
There are honest scientific debates here worth acknowledging. Some researchers question whether true gamma oscillations are even being entrained by sensory stimulation, and note limited propagation of the effect beyond the visual cortex, with limited engagement in key regions like hippocampal CA1. That’s a real concern. But the clinical signals are intriguing enough that several large trials are proceeding. Cognito Therapeutics, whose Spectris device uses a similar flicker approach, presented data at AAIC 2025 showing neuroprotective potential, and Sinaptica Therapeutics reported that personalized, non-invasive brain stimulation slowed cognitive decline by 44 percent in a Phase 2 trial for mild-to-moderate Alzheimer’s.
That number — 44 percent — deserves a moment. If that holds up in larger trials, it’s not a footnote. It’s a headline.
Deep brain stimulation: the invasive bet paying off slowly
If flickering lights are the gentlest intervention in this space, deep brain stimulation (DBS) is the most aggressive. It involves surgically implanting electrodes into specific brain structures and delivering continuous or patterned electrical stimulation. It’s already an established therapy for Parkinson’s disease and treatment-resistant depression. Alzheimer’s is a much harder target, but the early data is compelling enough that trials keep advancing. 🔬
The main target sites researchers have focused on include:
The fornix, the primary output pathway connecting the hippocampus to memory-related circuits
The nucleus basalis of Meynert (NBM), a cholinergic structure whose degeneration is one of the earliest events in Alzheimer’s pathology
The entorhinal cortex, the gateway through which sensory information enters the hippocampus
In the ADvance trial examining fornix DBS, patients who received the implant showed stable hippocampal volumes at the 12-month mark — a striking finding given that the typical Alzheimer’s comparison group loses roughly 5% of hippocampal volume per year. No atrophy. In a disease defined by tissue loss, that counts as something.
At the Medical College of Georgia, neuroscientist David Blake started human trials in January 2026 targeting the nucleus basalis of Meynert. His team’s protocol involves 50 minutes of DBS per day, delivered remotely by the patient or caregiver, using a 10-seconds-on, 40-seconds-off interval pattern Blake describes as “a little interval workout for the brain.” In animal studies, every single animal showed cognitive improvement without fail. That kind of result in a mouse or primate model doesn’t guarantee human translation, but it demands follow-through.
A 2025 clinical study in the journal CNS Neuroscience & Therapeutics compared fornix-DBS and NBM-DBS in patients with severe Alzheimer’s and found both improved cognitive function and quality of life, with NBM-DBS showing particular advantages in neuropsychiatric symptom management.
The question nobody can fully answer yet is long-term durability. Short-term cognitive enhancement is appearing across multiple DBS targets. Whether that enhancement holds at two years — and prevents the slide into total dependence — is what the current generation of trials is trying to find out.
The molecular wildcards: GABA receptors and nasal sprays
Not everything happening at the frontier requires electrodes or infusion suites. Some of the most intriguing work is molecular, and a couple of approaches are genuinely surprising in their simplicity. 💊
Researchers at the Centre for Addiction and Mental Health (CAMH) in Toronto, led by Etienne Sibille, identified a compound that targets alpha-5 GABA-A receptors — a subtype of the brain’s primary inhibitory receptors that regulate the balance between neural excitation and suppression. In Alzheimer’s, this balance tips badly wrong. The resulting spinoff company, Damona Pharmaceuticals, received FDA clearance for human clinical trials and planned to begin Phase 1 enrollment in early 2025. The mechanism is genuinely different from the amyloid-focused mainstream, which matters: one target being wrong doesn’t doom the other.
Meanwhile, the University of Texas Medical Branch published work in Science Translational Medicine on a different delivery system entirely. A nasal spray treatment showed promising results in clearing harmful tau protein buildup and improving cognitive functions in aged mice with neurodegenerative disease, with lead author Dr. Rakez Kayed noting that the approach “opens new avenues for non-invasive delivery of tau therapeutic antibodies directly to the brain.”
Then there’s the UCLA finding: a small molecule identified by Istvan Mody’s lab. When given to mice with Alzheimer’s disease, the treated animals — who previously couldn’t remember the route out of a maze — performed almost as well as healthy mice. Mody’s summary: “We may be able to restore cognitive function. That’s the ultimate hope.”
These aren’t the same mechanism, which is the point. Alzheimer’s isn’t one problem. It’s a cascade of failures — amyloid buildup, tau tangles, neuroinflammation, synaptic loss, oscillatory disruption, cholinergic degeneration — happening in sequence and in parallel. No single drug or device is likely to beat it. A combination will.
The bigger shift: diagnosis before damage
One thing rarely mentioned in mainstream Alzheimer’s coverage is that all of these treatments — drugs, DBS, gamma entrainment, molecular therapies — work far better the earlier you intervene. And the diagnostic picture has changed dramatically. 🧬
The FDA cleared the first blood test for diagnosing Alzheimer’s disease in May 2025, and the National Institute on Aging together with the Alzheimer’s Association released revised diagnostic criteria in 2024, formally defining the disease by its biology rather than its clinical symptoms. In practical terms: you no longer need a PET scan or a painful lumbar puncture to establish that Alzheimer’s pathology is present in the brain. A blood draw can do it.
This matters enormously for neurotech. Brain stimulation approaches like DBS and GENUS need neurons to work with. Anti-amyloid drugs need plaques that haven’t yet killed the circuitry downstream. The Lancet Commission’s 2024 updated report outlined 14 modifiable risk factors — including hearing loss, high LDL cholesterol, physical inactivity, and social isolation — that together could theoretically reduce global dementia cases by up to 45% if addressed systematically. The best neurotech in the world won’t save a brain that waited too long.
The opportunity, right now, is combining these tools: catch the disease early with a blood test, slow the amyloid cascade with lecanemab or donanemab, restore oscillatory function with non-invasive gamma entrainment, and use DBS or targeted molecular therapies in patients where the disease has already progressed further. As of early 2026, nearly 200 clinical trials are underway assessing more than 150 novel drugs, with an increasingly diverse range of biological targets well beyond amyloid.
We’ve covered the broader neurotech acceleration in 6 Signals That Neurotech Is Reaching a Tipping Point, and the Alzheimer’s story is one of the most vivid illustrations: investment, regulatory momentum, biological insight, and device innovation all moving in the same direction at the same time.
The honest question now isn’t whether the field is making progress. It clearly is. The question is whether the pace of clinical translation — notoriously slow, notoriously expensive — can keep up with the 55 million people worldwide currently living with dementia. What do you think needs to happen to close that gap?