Brain Stimulation

A brief history of brain stimulation research with a breakdown of present opportunities

The Future is Now

While the idea of brain stimulation might sound futuristic, especially when talking about automatic stimulation via brain implants, research on its means has a long history and contemporary applications are not uncommon.

Deep brain stimulation, the neurosurgical procedure of implanting brain-stimulating electrodes deep in select brain structures, has, for example, been in active use to treat Parkinson’s disease since 1997, reducing its negative symptoms up to 60% (Dallapiazza et al., 2018).

An even more common brain stimulation device in use today are hearing implants. Over 300k patients people have (re)gained the ability to hear thanks to electrode implants in their cochlea. Normally, the cochlea generates an auditory nerve impulse through audio vibrations picked up by tiny hairs on its inside, but if that fails, direct electrical stimulation of the nerve will suffice. This was first realised by Volta (also the inventor of the battery) in 1790, when he stuck metal rods into his ears to run electricity through his head and, to his delight, noted that this generated an audible crackling sound. Science! While it took another 167 years until this insight led to the partial restoration of the hearing ability of a deaf patient (Djourno and Eyries, 1957), it was only three years after (in 1960) that the first implantable cochlear stimulation device was built.

Since then, beyond electrical stimulation of the auditory nerve to restore hearing, advances have been made in the stimulation of the spinal cord to alleviate pain, of the vagus nerve to diminish depression, and even of the retina, choroid and sclera of the eye to restore vision, with currently existing technology being able to even give some degree of sight to the blind (Klauke et al., 2011; Jang et al., 2019).

In this article, I’ll give a brief history of the past of brain stimulation research, explain how modern neurosurgery is done, and how non-invasive means of brain stimulation might, in many cases, be a cheaper, more easily accessible alternative for treatment of neuropathologies or – potentially – as a means of brain augmentation.

With A Dubious Past

In a way, the idea that the brain could be meaningfully stimulated by electrical energy predates Volta’s findings by another 10 years, when Galvani and his assistants found that the limbs of dead frogs would move when their exposed nerves were shocked with a sufficiently strong electrical pulse. Galvani’s nephew Aldini then picked up his uncle’s torch in 1802 and took it one big step further, by stimulating the left cortical hemisphere of a recently decapitated criminal to evoke responses. From this point on, things moved fast.

Rolando, in 1829, was the first to experiment with more precise brain stimulation. Specifically, he sent electrical pulses to the cerebrum and cerebellum in dead pigs, in order to observe the effects and deduce wiring and function of stimulated brain areas. It was only a matter of time before, instead of experimenting with dead animals and humans, researchers would do so with live specimen.

Most interesting for our purposes are the groundbreaking – also on moral grounds – experiments by Roberts Bartholow, who was the first to give a detailed record of cortical stimulation in awake humans. He funneled insulated electrical wires through the protective layer of the brain of his test subject, and proceeded to apply an electric current, which elicited corresponding motor responses. While undoubtedly interesting to him and his peers (less so to us know, as we know so much more about the brain at this point), it also proved deadly for his test subject after three days of increasingly strong stimulation.

In 1888, Charles Mills produced the first map of cortical localisations of various brain functions assembled in similar (but mostly less morally dubious ways) from a number of sources. This was a milestone for neurosurgery in particular, which could now use maps like these to guide surgery. At around the same time, it also became common practice to electrically stimulate the exposed parts of the patient’s brain in close proximity to the surgical site, in order to avoid accidentally damaging vital parts of the brain structure. This is a practice which is still in use today and the reason why, in most cases, the patient has to be fully conscious during neurosurgery, as to be able to communicate the effects of said electrical stimulation as an additional safety procedure.

Another milestone in brain stimulation research was reached when, in 1963, Delgado invented the first fully wireless, implantable brain stimulation device which he dubbed the “stimoceiver”, as it could be activated remotely by radio waves. Knowing how to play media attention, he, to demonstrate its effectiveness, implanted it inside the brain of a bull to then challenge it to a bullfight in front of rolling cameras. Whenever the bull would charge at him, Delgado would press the button on his remote, activating the stimoceiver which then disrupted activity in the bull’s caudate nucleus. As this part of the brain plays a vital role in the control of voluntary movement, its disruption made the bull stop in its tracks – and severely confused it in the process. With the spotlight on him, Delgado, after testing the safety of stimoceivers with both macaques and human participants, then promoted the use of his device as a more humane alternative to the use of lobotomies.

Modern Neurosurgery

The examples of brain stimulation given so far require for the implanting of stimulating electrodes. The required brain surgery can be extremely demanding, especially if the modulation is aimed at deeper structures of the brain.

In general, any such surgical intervention starts with the creation of brain scans (usually MRI, see my article on Brain Imaging for details) to give a personal map of the individual brain structure of the patient. Then, the patients’ hair is shaved and a metal frame is attached to the head which fixes it in place. Intravenous antibiotics are injected, and a local anesthetic is applied directly to the scalp. Then, after making a small incision, a power drill is used to reach the outermost layer of the protective barrier between the skull and the cortex. Here, a cross-shaped cut is made to allow for the insertion of a small tube into the brain. No anesthetic is necessary here, as the brain does not contain any pain receptors. With the help of the thin tube, the electrode can then be pushed to the target region before its electrical leads are fixed to the inside of the skull. Once the surgery is complete, the electrode will be able to stimulate the affected region on demand. However, regular maintenance or battery replacement will require occasional additional surgeries – usually every few years.

Non-invasive Brain Stimulation

Brain stimulation implants will, for the foreseeable future, be mostly restricted to the most severe cases within the medical or clinical domain. For other applications, non-invasive brain stimulation might be both more easily accessible, much cheaper, and hold fewer risks.

An example of this is simple electrical stimulation of the scalp with either direct or alternating current. While embraced by some researchers and hobbyist biohackers, this approach has limitations due to a low spatial precision in stimulation. After all, electrical currents have a habit of spreading rather uncontrollably. On a more fundamental level, Vöröslakos et al. (2018) demonstrated that the voltage most generally applied by other researchers is attenuated by the skull and surrounding soft tissue to such a degree that the current actually reaching the brain is too minute to be able to modulate neural activity in the first place. Still, with a sufficiently strong current (>4.5 mA) this issue can be overcome. What remains, is the problem of a lack in spatial specificity of stimulation and, even more glaringly, the inability of surface-level electrical stimulation to specifically – and only – modulate sub-cortical activity. One promising technique to address both of these issues uses temporally interfering electric fields to stimulate more than just surface cortical areas and do so with a greater degree of precision than classical noninvasive electrical current stimulation. Specifically, Grossman et al. (2017) devised a method of applying electric fields to multiple points of the scalp which all oscillate at slightly differing high frequencies thought to not impact neural activity. When the high-frequency fields meet at their target location however, their modulation results in an electric field envelope that is driven by their difference frequency and hence slow and – given sufficiently strong stimulation – powerful enough to entrain and modulate groups of neurons.

In this example it is again electrical stimulation that is used to stimulate brain activity. This is an obvious approach, as the most overt form of neural communication functions via the propagation of electrical impulses. Every electrical field comes with a magnetic field and vice versa, however. For example, we can generate a magnetic field by running a current through coiled wire (which, in this context, would be referred to as an inductor). Even though magnetic fields are much weaker than electric ones, given big enough coils they are strong enough to influence the current inside groups of neurons – even when the coil is not implanted but just placed against the scalp. This means that we can also induce electrical activity in the brain through the use of magnetic fields. This approach is called transcranial magnetic stimulation (TMS) and it’s among the best validated forms of noninvasive neuromodulation. During my PhD I too had the opportunity to run experiments with it, and was able to experience how magnetic stimulation of my motor cortex caused involuntary movements of my arm (sometimes without conscious awareness of said movement), or how TMS pulses directed at my visual cortex brought about the illusion of light flashes when my eyes were closed. One major reason for the popularity of TMS among researchers is how well evidenced the mechanisms are with which it affects neural activity. Another is its deep reach in the brain and its – compared with electrical stimulation – high degree of precision in stimulating groups of neurons, albeit it’s still a sledgehammer method when contrasted with invasive forms of brain stimulation. After all, aside from unwanted propagation of activity through non-target areas of the brain as a consequence of the traveling path of the magnetic pulse, each pulse will trigger thousands of neurons in the periphery of the target region, which itself is too broad an area for any precise modulation.

As both electrical stimulation and even TMS are lacking in precision, to me the most promising form of non-invasive brain stimulation comes in the form of focused ultrasound stimulation (fUS). It’s a fairly new protocol that applies ultrasound transducers at a number of scalp locations, and uses the constructive interference of groups of sound waves at the brain target location to boost their amplitude. As a result, localised vibrations caused by the resulting sound waves then affect the activity of ion channels, boosting neural excitability in the region. As the propagation of sound waves has a narrower spread than either electric currents or magnetic pulses, fUS has a significantly higher spatial accuracy in brain stimulation than either electrical interference stimulation or TMS.

Rhythmic Capture

All forms of stimulation described so far are generally applied in bursts. As networks of neurons fire in rhythmic intervals, with the frequency of their firing constituting a form of pulse-rate coding, bursts of stimulation are sometimes applied rhythmically as well: When the rhythmicity of the stimulation applied falls within range of the firing frequency of a network of neurons, their oscillation falls in sync – is entrained – by the exogenous driving signal. In other words, rhythmic stimulation, rather than stimulating individual neurons, is able to selectively modulate neural networks which are characterised by a distinct firing rate.

Modulating or initiating any oscillator endogenous to a system via exogenous forcing is referred to as entrainment. Flexibility of the oscillating system is a necessary prerequisite for this. Entrainment has functional relevance, as it allows for adaptivity of the oscillator to its environment. Consequently, they are a core feature of neural oscillators, which can be exploited by using rhythmic brain stimulation techniques such as photic stimulation (stroboscopic flashes at specific frequencies).

This exploits the fact that the rhythmic temporal architecture of the brain makes it prone to be captured by regularities in input. An example of this is how, when listening to any audio input, activity in your auditory cortex will synchronise to the audio rhythmicities. Similarly, visual cortex activity quickly falls in lockstep with regularities in visual stimuli. Functionally, this makes the brain more sensitive to patterns which allow for effective pre-allocation of limited cognitive resources for fast (automatic) responses to somewhat familiar situations. For our purposes, however, it gives us an easy hack to entrain the brains’ rhythmic structure in order to regulate the amplitude, phase and frequency of its local and global neural networks, which holds promise to normalise or enhance brain function.

Why Stimulate the Brain

Brain stimulation promises us a new degree of autonomy over our own brain. If we can modulate our own neural activity, this could provide a path to treating brain disorders, reverse neurodegenerative effects of aging, disease or injury, and even uplift our potential by adding to the brains’ computational power. The potential pitfalls are huge due to an incomplete understanding of the brain and sometimes lacking precision of brain stimulation solutions. In addition, there are high costs associated with neurosurgery in particular. Yet, especially in medical and clinical fields, success stories abound, and we’ve successfully restored a degree of hearing in the deaf, brought limited vision back to the blind and partially restored the movement of people suffering from Parkinson’s – just to name a few. What started as a rather crude affair is becoming an increasingly sophisticated and precise discipline – albeit the challenges to be overcome are numerous still.


Dallapiazza, R.F., De Vloo, P., Fomenko, A., Lee, D.J., Hamani, C., Munhoz, R.P., Hodaie, M., Lozano, A.M., Fasano, A. and Kalia, S.K., 2018. Considerations for Patient and Target Selection in Deep Brain Stimulation Surgery for Parkinson’s Disease. In Parkinson’s Disease: Pathogenesis and Clinical Aspects [Internet]. Codon Publications.

Djourno, A. and Eyries, C., 1957. Auditory prosthesis by means of a distant electrical stimulation of the sensory nerve with the use of an indwelt coiling. La Presse Médicale65(63), pp.1417-1417.

Grossman, N., Bono, D., Dedic, N., Kodandaramaiah, S.B., Rudenko, A., Suk, H.J., Cassara, A.M., Neufeld, E., Kuster, N., Tsai, L.H. and Pascual-Leone, A., 2017. Noninvasive deep brain stimulation via temporally interfering electric fields. Cell169(6), pp.1029-1041.

Jang, J., Kim, H., Song, Y.M. and Park, J.U., 2019. Implantation of electronic visual prosthesis for blindness restoration. Optical Materials Express9(10), pp.3878-3894.

Klauke, S., Goertz, M., Rein, S., Hoehl, D., Thomas, U., Eckhorn, R., Bremmer, F. and Wachtler, T., 2011. Stimulation with a wireless intraocular epiretinal implant elicits visual percepts in blind humans. Investigative ophthalmology & visual science52(1), pp.449-455.

Vöröslakos, M., Takeuchi, Y., Brinyiczki, K., Zombori, T., Oliva, A., Fernández-Ruiz, A., Kozák, G., Kincses, Z.T., Iványi, B., Buzsáki, G. and Berényi, A., 2018. Direct effects of transcranial electric stimulation on brain circuits in rats and humans. Nature communications9(1), pp.1-17.