There are as many as 86 billion neurons in the human brain. They make trillions of synaptic connections, with each neuron connecting with other neurons to create a complex network.
It is through synaptic signaling that neurons communicate with each other at synapses, which are junctions where neurons can excite or inhibit other neurons.
A synaptic connection, meanwhile, is the junction between neurons where they communicate with each other using electrical and chemical signals. The mapping of synaptic connections is a process of identifying just how neurons are connected to each other and how strong these connections are. The structure of these connections is believed to determine brain function, making this process extremely important.
However, mapping and characterization of synaptic connections is an open challenge. Currently, we are limited to the mapping of only a few hundred synaptic connections. However, the latest study from Harvard researchers has expanded it to over 70,000 with accuracy, making it a major breakthrough in the field of neuroscience, particularly neuronal recording.
The mapping and cataloging of this scale were achieved from about 2,000 rat neurons with the help of a silicon chip with microhole electrodes that can detect small but notable synaptic signals simultaneously.
Such an advancement can help us create a detailed map of synaptic connections and allow us to better understand how neurons connect and communicate.
The study titled “Synaptic connectivity mapping among thousands of neurons via parallelized intracellular recording with a microhole electrode array” was published1 in Nature Biomedical Engineering this month.
The parallelization of neuronal intracellular recording at a large scale, according to the study, is a big challenge that must be overcome as it allows for measuring synaptic signals across a neuronal network, which in turn allows for the mapping as well as characterization of synaptic connections.
Scaling Up Neuronal Recording with Advanced Chip Innovation
For the visualization of synaptic structures, electron microscopy (EM) has been found to be highly effective, but it can’t measure the strength of neuron connections. This limits the technique’s ability to explain the functionality of neuronal networks.
This is where the patch-clamp recording comes in, offering a more precise technique to study synaptic activity as it can pierce through individual neurons and capture even weak signals with high sensitivity.
Scientists Erwin Neher and Bert Sakmann received the 1991 Nobel Prize in Physiology or Medicine for the patch-clamp technique. The experimental measuring device proved the existence and function of ion channels in driving important tasks in living cells like muscle contraction and nerve impulse transmission. The technique also allowed key features of the channels to be defined with fine precision.
While this method can help researchers identify and measure connection strength, scientists have been struggling to record intracellular signals from more than a few neurons at a time until now.
Led by co-senior author Donhee Ham, the Professor of Engineering and Applied Sciences at Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS), the researchers from this department, as well as Harvard’s Department of Chemistry and Chemical Biology, built an advanced chip.
More precisely, the team has built an array of 4,096 microhole electrodes on a chip. This device conducted the massively parallel intracellular recording of about 2,000 rat neurons.
This provided researchers with abundant data on synaptic signals, from which they extracted more than 70,000 synaptic connections.
The work actually builds upon the breakthrough device that the team developed a few years ago. At the time, an array of 4,096 vertical nanoneedle electrodes was sticking out of a silicon chip of the same integrated circuit design.
The electronic chip featured a dense array of vertically-standing, platinum powder-coated (to improve signal passing ability) nanometer-scale electrodes on its surface.
On that device, a neuron could wrap around a needle to allow intracellular recording, parallelized through a large number of electrodes. At the time, the team was able to record more than 1,700 rat neurons and map over 300 synaptic connections.
The numbers achieved were record-breaking, but the team felt that they could improve them even more, hence, the latest study.
Click here to learn if memristors are key for mimicking human brain.
Revolutionizing Intracellular Recording for Connectivity Mapping
Supported by the Samsung Advanced Institute of Technology of Samsung Electronics, the team designed and fabricated the platinum/platinum-black microhole electrode array on the complementary metal-oxide semiconductor (CMOS) chip.
The array was operated to gently open up cells by injecting small currents through the electrodes to parallelize their intracellular recording.
The integrated electronics in the silicon chip, according to study co-lead author and former postdoc researcher Woo-Bin Jung, who’s currently a faculty member at Pohang University of Science and Technology in South Korea, “plays as equally an important role as the microhole electrode, providing gentle currents in an elaborate way to obtain intracellular access, and recording at the same time the intracellular signals.”
The microhole design used here, meanwhile, is similar to the patch-clamp electrode. These microhole electrodes have two main benefits:
- They are much easier to fabricate
- They couple to neurons’ interiors better than the vertical nanoneedle electrodes the team used previously
“This accessibility is another important feature of our work.”
– Jun Wang, Postdoctoral Researcher and Co-lead Author
With this new design, the team was able to achieve great results. On average, 90% of microhole electrodes (over 3,600 out of 4,096) were intracellularly integrated, which helped the team extract 70,000 plausible synaptic connections from them, much more than the 300 with the nanoneedle electrode array.
The team didn’t just get quantitative results but also achieved quality in their recording data, which allowed them to classify synaptic connections per their strengths and characteristics.
Synaptic connections were cataloged into electrical synaptic connections and strong/eventful excitatory, weak/uneventful excitatory, and inhibitory chemical synaptic connections, with an estimated 5% overall error rate.
According to the study, the mapping of synaptic connectivity at such an extensive level, as well as the ability to identify synaptic connections, is a big move toward the functional connectivity mapping of large-scale neuronal networks.
In their previous version, the researchers used the high-performance chip to calculate the effects the drugs have on synaptic connections. At the time, the team stated that they were building a wafer-scale system for drug testing for neurological disorders like addiction, autism, Alzheimer’s disease, Parkinson’s disease, and schizophrenia.
The biological synaptic network mapping can also provide a new strategy for machine intelligence to build next-generation artificial neural networks and neuromorphic processors.
Now that the team has gained insights into synaptic connections from the massive amount of data they got from parallel intracellular recording, they are “working toward a newer design that can be deployed in a live brain.”
Click here to learn how AI & neuroscience are combinedly decoding human brain.
Advances in Brain Mapping & Neurotechnology
Studies like these are crucial in deepening our understanding of the brain’s intricate synaptic connections, revealing not just the complexity of neuronal interactions but also providing a foundation for advanced research in brain mapping, one of many exciting developments shaping the future of brain research.
Brain, after all, is an interesting but complex structure, making it a field of growing interest among researchers and institutions. Interestingly, the more we study and get to know our brains, the more complex they become. As Jeff Lichtman, a professor of molecular and cellular biology at Harvard, noted, simply describing nature has revealed:
“[That the brain is] more complicated than we thought.”
For over a decade now, Lichtman has been accessing a specific (one cubic millimeter) area of the cerebral cortex to create a detailed map of connections in the human brain. Late last year, Harvard Magazine reported their analysis to discover that glial cells that support and protect neurons dominate them 2-to-1. However, an explanation for the same has yet to be found.
He and his colleagues also observed a few very strong connections with many synapses for every cell to allow the information to flow through the brain very quickly, which, according to them, “maybe… what memories are.”
For this, they used tissue from a woman’s temporal lobe and sliced it into over 5,000 sections, with each one scanned with a purpose-built multibeam electron microscope at a very high resolution.
Google researchers then built new neural networking methods to identify items in each of these slices and combined them to create a 3D space. The seed-sized sample they imaged contained 57,000 cells, 230 millimeters of blood vessels, and 150 million synapses.
Lichtman next aims to map the entire brain of a mouse, as doing so with the human brain is not yet possible. Meanwhile, a team of scientists from Princeton mapped an adult fruit fly’s brain.
Previously, researchers have mapped the brain of a larval fruit fly with 3,000 neurons, but the adult fruit fly has almost 140,000 neurons and tens of millions of synapses connecting them. The map was built using 21 million images of the fruit fly brain and an AI model. With the help of this blueprint, the researchers aim to provide answers to which neurons are responsible for which behaviors.
The human brain, as John Ngai, director of the US National Institutes of Health’s BRAIN Initiative, stated, “is more powerful than any human-made computer,” in a lot of respects, “yet for the most part, we still do not understand its underlying logic.”
And if we don’t have “a detailed understanding of how neurons connect with one another, we won’t have a basic understanding of what goes right in a healthy brain or what goes wrong in disease.”
This is why there’s a growing focus on trying to map out our brains. Scientists at The Research Institute of the McGill University Health Centre actually implemented a method called “optomapping” this month to dramatically accelerate the study of neurons.
The team has already tested more than 30,000 candidate connections and characterized about 1,800 synapses in the mouse brain’s visual cortex.
Key discoveries made using optomapping include unique connection patterns onto different types of inhibitory neurons, which challenge what we have known about the information flow process in the brain’s cortical layers. They also demonstrated the careful balancing of the brain’s communication patterns, with some connections amplifying signals and others calming them.
Furthermore, they observed an imbalance in how different neuron types are activated, which can help in conditions like autism and epilepsy. By pinpointing changes, Optomapping could lead to more targeted and effective treatments.
Interestingly, this month, the tech giant Meta Platforms (META -1.62%) announced that it had developed a device that lets people produce text just by thought. For this, a brain scanner and an AI model called Brain2Qwerty were used to decode language from the brain without surgery.
However, the chances of the product going commercial are too low due to several constraints. For instance, the magnetoencephalography scanner used for detecting magnetic signals is extremely large, weighing about half a ton, and expensive, costing $2 million. It also works only in a shielded room to dampen the magnetic field of Earth.
But this can certainly help advance our understanding of the human brain and communication.
“Trying to understand the precise architecture or principles of the human brain could be a way to inform the development of machine intelligence. That’s the path.”
– Meta’s Brain & AI team lead, Jean-Rémi King, told Tech Review.
The findings show that the system can detect what keys a “skilled” typist would hit with an accuracy as high as 80%. While not flawless, the researchers said it’s accurate enough to construct full sentences from brain signals.
These developments coming from a $1.76 trillion market cap company, which is primarily involved in social media and AI, may seem surprising to some, but Meta has been involved in neurotechnology research for many years now.
Also, unlike companies involved in pure neurotechnology play, Meta is a profitable one, which means it can invest in related research without worrying about resources.
Meta Platforms, Inc. (META -1.62%)
For context, the company recently reported its Q4 results, which showed a 21% YoY increase in revenue at $48.39 billion. After all, an average of 3.35 billion daily active people (DAP) used its products. During this period, Meta recorded $25.02 bln in costs and expenses and $14.84 bln in capital expenditures, while its free cash flow was $13.15 billion, and cash, cash equivalents, and marketable securities were $77.81 billion. It has also announced an investment of about $60-$65 bln in AI this year.
Companies Advancing This Space
Now, let’s take a look at prominent publicly listed companies that are making significant advancements in the field.
1. Neuronetics, Inc. (STIM -13.96%)
This medical technology company develops products for patients who suffer from neurohealth disorders. Its NeuroStar Advanced Therapy System is a non-invasive treatment that uses transcranial magnetic stimulation (TMS) to stimulate areas in the brain that are associated with mood. It is used to treat adult patients with major depressive disorder (MDD).
With a market cap of $267 million, STIM shares are trading at $4.80 as of this writing, up a whopping 198.14% YTD. Its EPS (TTM) is -1.22, and its P/E (TTM) is -3.94.
Neuronetics, Inc. (STIM -13.96%)
For Q3 2024, Neuronetics reported $18.5 million in revenue, a net loss of $13.3 million, and a decline in cash position to $20.9 million from $59.7 mln at the end of 2023. During this period, the company shipped 48 NeuroStar systems in the US, and its revenue reached $4.1 million, while treatment sessions grew by 2%.
Neuronetics also acquired Greenbrook TMS, which is expected to accelerate its path to profitability, as per CEO Keith J. Sullivan.
According to preliminary Q4 results, Neuronetics’ revenue for Q4 is estimated at $22.1 million and $74.5 million for the full 2024. However, it has announced better expectations for this year, with total revenue between $145 million and $155 million and positive cash flow in the third quarter.
#2. NeuroOne Medical Technologies Corporation (NMTC +2.68%)
This medical technology company develops minimally invasive solutions for brain stimulation, EEG recording, and ablation solutions for those suffering from neurological disorders like epilepsy, dystonia, tremors, Parkinson’s disease, and chronic pain.
Its products include Evo cortical electrode technology for recording, monitoring, and stimulating brain tissue.
Then there’s the OneRF ablation system for creating “radiofrequency lesions in nervous tissue for functional neurosurgical procedures.” NeuroOne has actually achieved FDA clearance for its OneRF Ablation System, making it the first and only one to receive it for both electrical activity reporting and nervous tissue ablation.
NeuroOne Medical Technologies Corporation (NMTC +2.68%)
With a market cap of $34.56 million, NMTC shares are currently trading at $1.12, up 35.61% YTD. Its EPS (TTM) is -0.27, and P/E (TTM) is -4.21.
In the first quarter of fiscal 2025, the company reported $3.3 million in product revenue and $1.9 million in product gross profit while operating expenses decreased. As of December 31, 2024, it had cash and cash equivalents of $1.1 million, lower than the $1.5 million it had as of September 30, 2024.
Conclusion
Breakthroughs in brain-related research are critical for the advancement of humans. After all, it is our most important organ, which controls nearly every function, from thought, memory, and emotion to movement, vision, and hearing. So, having a deeper understanding of the brain is a necessity, and the latest research making advances in large-scale synaptic connectivity mapping marks a significant step in decoding just how neurons interact and communicate.
As these advancements continue to refine, they can pave the way for deeper insights into the brain’s functionality, assist in new therapeutic strategies for complex conditions, and revolutionize neurological research.
Click here for a list of top blue chip pharmaceutical companies.
Study Reference:
1. Wang, J., Jung, W. B., Gertner, R. S., et al. (2025). Synaptic connectivity mapping among thousands of neurons via parallelized intracellular recording with a microhole electrode array. Nature Biomedical Engineering. https://doi.org/10.1038/s41551-025-01352-5