Recently developed non-invasive deep brain stimulation methods have sufficient focal specificity to target deep brain structures. These techniques show particular promise as treatment strategies for neuropsychiatric disorders in which deep brain structures have critical roles in pathophysiology or in mediating recovery.

Deep brain structures such as the basal ganglia, thalamus and hippocampus are core brain areas in the pathophysiology of neurological and psychiatric disorders such as dementia, Parkinson disease, anxiety disorders, stroke and traumatic brain injury (TBI). Furthermore, these structures are involved in transdiagnostic symptoms, such as memory deficits, apathy and fatigue, that appear in multiple neurological and psychiatric disorders despite different underlying pathophysiologies. Thus, interventional neuromodulation-based treatment strategies that target these brain areas have promising and diverse applications. However, because of their locations deep in the brain, these regions were, until recently, accessible only by invasive deep brain stimulation (DBS) or lesioning approaches, for example, with focused ultrasound. Direct stimulation of these targets with established non-invasive neuromodulation methods, such as transcranial magnetic stimulation (TMS) or transcranial electrical stimulation (tES), was not feasible owing to the limited depth–focality trade-off of the techniques, meaning that if deep brain structures were targeted, overlying cortical areas would be strongly co-stimulated. This factor strongly limited progress in translational research into non-pharmacological neuromodulation techniques and their ability to leverage the core functional role of deep brain structures in transdiagnostic symptoms and neuropsychiatric disorders.

In the past few years, an innovative approach developed in animal models1,2 to non-invasively neuromodulate deep brain structures using tES has been successfully applied in, to our knowledge, the first human proof-of-concept studies3,4,5. The technique, called transcranial temporal interference electrical stimulation (tTIS), involves the application of two sources of high-frequency electrical current in the kilohertz range (for example, 2,000 Hz), to which neurons do not react (see ref. 1 for details); then, when a slight difference in frequencies is introduced to the two currents — such as 2,010 Hz versus 2,000 Hz — a beating envelope is created at the delta frequency within the physiological range (here, 10 Hz). The resulting envelope amplitude peak can be steered towards locations deep within the brain. Thus, this approach can overcome the problem of depth–focality trade-off to enable neuromodulation of deep brain structures3,4,5.

In one of the first human trials of tTIS, in healthy individuals, theta-burst stimulation that induces plasticity was applied to the striatum, a core region in the process of motor learning and re-learning after brain damage3. The results showed a strong enhancement of the learning process compared with a placebo condition, especially in older participants aged 60–90 years3. By contrast, clinical trials of other stimulation techniques that target the motor cortex to enhance rehabilitative re-learning, such as the NETS Trial6, have shown limited effect sizes or no effects. Thus, tTIS seems to offer a better prospect for the use of non-invasive brain stimulation to enhance motor re-learning during rehabilitation compared with stimulation of the cortex, which might be involved in only the initial part of the re-learning process7.

In another study, Vassiliadis et al.5 used tTIS to target striatal mechanisms of reinforcement learning in healthy individuals. The technique was shown to disrupt ongoing oscillatory activity in a frequency-specific manner5, which presents the opportunity to interfere with pathological oscillatory activity, such as in essential tremor or Parkinson disease. Furthermore, Violante et al. demonstrated that tTIS of the hippocampus in the theta range can modulate associative memory in healthy individuals aged 18–35 years, probably via entrainment of the target frequency, especially when stimulation is steered to the task-critical subregion of the hippocampus4.

“the technique can be applied in the three main conceptual domains of neuromodulation”

Together, the initial human trials of tTIS demonstrate that the technique can be applied in the three main conceptual domains of neuromodulation: enhancement of plastic properties3, entrainment of oscillatory activity4 and disruption of ongoing physiological or pathological oscillations5, with a good safety and blinding profile8,9. Therefore, tTIS has great potential for applications in neurological and psychiatric disorders without the requirement for a surgical procedure with its related risks and costs. This non-invasive approach might also have uses that are complementary to invasive DBS. For example, tTIS could provide predictive testing for response likelihood before an invasive treatment is implemented, for example, in patients with Parkinson disease or essential tremor. Furthermore, the non-invasive nature of tTIS with its online steering capabilities could enable a dynamic multi-target approach to stimulate further symptom networks in order to target symptoms that do not respond well to invasive DBS or even worsen after implantation, such as speech or cognitive impairments, or postural instability10. Last, tTIS could be used for interventions that do not need permanent neuromodulation, such as motor or cognitive rehabilitation after stroke or TBI, and to enhance resilience against cognitive decline concurrent to cognitive training.

Pilot testing of tTIS in clinical patients included two people with Parkinson disease and one with essential tremor in whom a stimulation frequency resembling classical DBS protocols (130 Hz) was applied to disrupt pathological oscillations2. In this case study, stimulation of the subthalamic nucleus led to improvements in resting tremor amplitude and continuity. Although promising, these results should be interpreted with caution owing to the small number of participants and incomplete control conditions. Of note, the applied stimulation protocol differs considerably from conventional protocols for invasive DBS: tTIS is applied in a sinusoidal rather than pulsed pattern, and at a sub-threshold intensity that modulates the impact of ongoing neuronal activity, rather than at a supra-threshold intensity that generates paced action potentials. As a result, the different types of neurostimulation operate via different mechanisms of action. Further proof-of-concept evidence was presented at the European Academy of Neurology Congress 2024 in Helsinki, and showed that the application of plasticity-inducing tTIS by means of theta-burst stimulation can enhance motor learning and associative memory when applied to the striatum or hippocampus, respectively, in individuals with brain damage from TBI (F.C.H., unpublished work). These results highlight the potential of deep tTIS for the treatment of neurological and psychiatric disorders.

“plasticity-inducing tTIS by means of theta-burst stimulation can enhance motor learning and associative memory”

So far, the demonstrated effects of tTIS have been short lasting and unlikely to be clinically relevant, as expected for the single-session interventions used in the described trials. Future clinical work with repeated sessions is needed to demonstrate longer lasting, clinically relevant improvements. In addition, tTIS electrode placement is guided by simulations; however, currently the models are based on healthy brains. Thus, another crucial area of future research is to derive electrical field simulations on an individual level from people with brain damage caused by neurodegeneration or lesions. Furthermore, improvement in the mechanistic understanding of the intervention is needed, for example, in relation to which neuronal population in the striatum is modulated by tTIS. This knowledge will allow us to optimize the parameters of stimulation protocols, leading to a more focal, personalized or state-dependent and closed-loop stimulation.

To ensure the reliability and usefulness of tTIS as an adjuvant treatment and its translation into the clinic, these questions have to be successfully addressed. Moreover, the efficacy of this non-invasive DBS must be rigorously tested in well-defined but focused randomized clinical trials. A further goal will be to determine the potential of tTIS in conjunction with classical DBS, for example, to define responders or non-responders or to address symptoms that do not respond well to invasive DBS.