Definition, system components, and boundaries
A repetitive transcranial magnetic stimulator uses rapidly changing current in a coil to create pulsed magnetic fields. Those fields pass through intact scalp and skull and induce an electric field in brain tissue. Delivering pulses in trains or structured patterns can alter activity in neural circuits without an incision. A typical system includes a high-voltage charging and discharge unit, treatment coil, cooling hardware, articulated coil holder, operator console, trigger and emergency stop, treatment chair, targeting tools, and session-recording software. Some systems add electromyography, image-guided navigation, a powered positioning arm, or a purpose-built helmet coil.
This entry concerns systems that deliver repeated pulses for treatment. Single- and paired-pulse TMS are often used to examine motor conduction, cortical excitability, and circuit interactions. Conventional rTMS and intermittent or continuous theta-burst patterns are repetitive methods, but their internal timing and safety envelopes differ. Deep TMS uses a particular coil geometry to produce a broader and differently distributed field; it does not move an unchanged, sharply focused field harmlessly into deep brain tissue. Magnetic seizure therapy deliberately induces a generalized seizure under anesthesia. Electroconvulsive therapy, implanted deep-brain stimulation, and transcranial electrical stimulation use different energy, procedures, and risk controls.
The ability of a magnetic pulse to cross bone does not make its biological action restricted to the intended target. Induced electric field depends on coil geometry, waveform, head shape, cortical folding, tissue conductivity, coil-to-scalp distance, and coil angle. Scalp muscles and peripheral nerves may also be activated. Similar product names, enclosures, or frequency labels do not establish that two coils, targets, doses, or clinical effects are interchangeable.
Waveform, coil geometry, and complete dose specification
A reproducible prescription identifies the stimulator and coil model, monophasic or biphasic waveform, pulse width, coil geometry and orientation, repetition frequency, train duration, inter-train interval, trains and total pulses per session, output intensity, session schedule, complete course, and any accelerated or maintenance plan. Intensity is commonly expressed as a percentage of resting or active motor threshold. The threshold definition, tested muscle, measurement method, and date must accompany that percentage. A statement such as “100 percent intensity” alone has no portable meaning.
The geometric center of a coil is not invariably the point of maximal electric field. Figure-of-eight, circular, and multiwinding coils differ in focality, direction, and depth profile. Field magnitude falls rapidly as a coil moves away from the scalp. Hair, spacers, coil covers, tilt, and applied pressure can therefore alter exposure. A larger coil or a claim of greater depth commonly entails wider superficial exposure; a penetration distance in centimeters does not describe selectivity. Percentage of maximum stimulator output is not a unit of intracerebral electric field. Comparison between systems requires appropriate electromagnetic measurement or an adequately validated individualized field model.
Frequency is only one dimension of dose. The common description of low frequency as inhibitory and high frequency as excitatory is a population-level simplification. Baseline state, target, intensity, train structure, and prior activity can alter both magnitude and direction of response. Theta-burst protocols place pulses into short bursts that repeat at another frequency, and intermittent and continuous forms are not equivalent merely because their total pulse counts match. A change in frequency, intensity, train length, interval, or number of same-day sessions changes the protocol as a whole and requires a corresponding safety and evidence review.
Motor threshold, targeting, and delivery of a session
Motor threshold estimates individual excitability and system-to-head coupling from reproducible motor-evoked potentials or a defined visible motor response elicited over motor cortex. Electromyography quantifies response amplitude and successful trials, whereas visual observation is more vulnerable to posture and observer effects. Resting and active thresholds are different measures. Coil position, unintended muscle contraction, medication, sleep loss, and the threshold-search algorithm can all influence the result. A change of stimulator or coil, a clinically important medication change, or a substantial change during the course may require repeat measurement under the device labeling and clinical protocol.
A treatment target may be derived from a fixed scalp distance, a proportional rule, a standardized electroencephalography coordinate, individual structural imaging, or functional-connectivity navigation. Simple rules are repeatable but can miss the same cortical region in heads with different size and anatomy. Navigation improves geometric reproducibility, yet does not by itself show that the selected network is optimal for a particular patient. Before every session, staff should confirm identity, side, target, coil face and handle direction, stand-off distance, chair position, and head reference. Translational and rotational tolerances should be defined. Movement that breaks registration requires repositioning rather than simply returning the arm to an approximate location.
Treatment is usually a course of outpatient sessions, not a one-time immediate intervention. Before starting, the operator explains the clicking sound, tapping or muscle contraction, and the patient's stop signal. Intensity is brought to the prescribed level while tolerability is observed. Rapid repetitive discharge heats a coil, so validated cooling and temperature interlocks are part of the safety system. Overtemperature, cooling failure, an unusual odor, output error, or loose positioning calls for stopping and inspection. Accelerated schedules, multiple targets, and large research pulse totals should not be copied outside the evidence and supervision that define them.
Clinical evidence, indications, and regulatory interpretation
Evidence is most established for rTMS in major depressive disorder, with randomized trials and professional consensus supporting several prefrontal protocols. Outcome still depends on patient selection, prior treatment, the exact coil, and the authorized protocol. In some jurisdictions, particular systems and protocols also have indications related to obsessive-compulsive disorder, migraine, or assistance with smoking cessation. Authorization attaches to a model, indication, population, coil, parameters, concomitant care, and labeling. It does not mean every rTMS system is authorized for every such condition, and neither device registration, research use, nor similarity to a marketed system is approval of effectiveness.
Clinical assessment should use prespecified symptom scores, response and remission, function, quality of life, recurrence, and durability. A course spans repeated sessions. Lack of a dramatic sensation early in a course is not proof of failure, while brief alertness or relaxation is not a reason to increase dose. Depression with psychotic features, imminent suicide risk, mania, catatonia, or another acute emergency may require a different or faster intervention; rTMS must not delay crisis assessment. Continuation and maintenance schedules are not universal and should account for relapse history, prior response, medication, and psychotherapy.
Complete blinding is difficult because active stimulation produces sound, tapping, and facial muscle movement. Turning a coil away changes both sensation and induced field. A critical reading asks whether the sham matched auditory and somatic cues, whether participants and assessors guessed assignment, whether the primary outcome was registered in advance, whether attrition and concurrent treatments were balanced, and whether a result was driven by a small single-center sample or multiple comparisons. A change in neuroimaging, electroencephalography, or motor threshold is a mechanistic observation, not a substitute for an outcome that matters to patients.
Implant screening, adverse effects, and emergency readiness
Screening addresses metal or fragments near the head and neck, aneurysm clips, stents, cochlear implants, ocular or intracranial implants, and deep-brain, vagus-nerve, spinal-cord, or other stimulators. Pacemakers, defibrillators, medication pumps, leads, and wearable electronics add electromagnetic-compatibility and separation questions. Safety cannot be inferred from whether a metal is attracted to a magnet: a conductive loop can carry induced current, and an electronic device can malfunction. Staff should identify the exact implant model and manufacturer conditions and remove hearing aids, jewelry, phones, and other items not cleared for proximity to the coil.
An induced seizure is the principal serious acute risk, although it is uncommon during protocol-conforming clinical use. Prior seizures, brain injury or stroke, tumor, central nervous system disease, abnormal intracranial pressure, electrolyte disturbance, fever, sleep deprivation, withdrawal from alcohol or sedatives, and medicines that lower seizure threshold can change risk. Screening is not a simple permanent yes-or-no list; the specific protocol and specialist judgment matter. Personnel must distinguish seizure from syncope, prevent a fall, stop stimulation and move the coil away, time the event, and follow a rehearsed emergency and referral plan. Convulsing limbs should not be forcibly restrained.
Common adverse effects include application-site pain, headache, facial or scalp muscle twitching, neck discomfort, fatigue, and transient dizziness. Coil discharge is loud enough to create a hearing hazard. Patients and exposed staff need hearing protection appropriate to the device acoustic output and labeling, with reassessment if plugs fit poorly or tinnitus, ear pain, or hearing change occurs. Monitoring also covers worsening mood, suicidal thinking, agitation, mania or hypomania, cognitive change, and persistent neurologic symptoms. Pregnancy, childhood, epilepsy, and complex implants require additional product-specific assessment; reassuring experience in routine adults cannot simply be extrapolated.
Quality assurance, documentation, and longitudinal management
Acceptance testing, periodic maintenance, repair, and software updates should be followed by checks of output waveform and amplitude, timing, coil recognition, cooling, emergency stop, interlocks, and navigation accuracy. Cracks or heating in a cable, enclosure, or connector may precede an obvious output failure. Heavily used coils require tracking of pulse count and manufacturer service life. A replacement stimulator or coil cannot be assumed to deliver the same dose merely because the interface displays the same percentage. The treatment area also needs control of flammable material, noise, and unrelated electronics, while preserving access for resuscitation and transfer.
Each session record should identify device, coil, and software version; motor threshold and method; target and localization method; intensity, frequency, train length, interval, and total pulses; session number; positioning deviations; pauses; alarms; and adverse effects. Medication, substance use, sleep on the preceding night, and major health changes help interpret tolerance and threshold shifts. Standard clinical measures belong at defined time points and should be completed by qualified staff. Device completion rates and same-day impressions alone are inadequate, and failed or prematurely stopped sessions must not disappear from the effectiveness record.
Scheduling and cloud platforms may hold psychiatric symptoms, suicide-risk data, medication, brain images, and precise treatment logs. Access should be limited, transfers encrypted, and retention and export rules explicit. An algorithm that recommends a target, intensity, or course needs a stated training population, applicable population, human review, and safe fallback. rTMS remains a medical procedure that requires screening, trained supervision, and emergency capability. Its noninvasive appearance is not a basis for unsupervised home “brain fitness,” nor should consumer sleep scores automatically increase treatment dose.
Sleep and insomnia: limits of the evidence
Sleep studies of rTMS include primary or chronic insomnia, sleep symptoms accompanying depression, and post-stroke sleep disturbance, which are not one population. Earlier meta-analyses reported possible improvement on self-rated measures such as the Pittsburgh Sleep Quality Index. Polysomnographic total sleep time, efficiency, latency, and staging were not consistently improved. Targets, frequencies, intensities, session counts, medication, and sham methods varied widely. More recent pooled analyses continue to report favorable signals, but they do not erase heterogeneity, reporting bias, or limitations in certainty. Sleep improvement that follows improvement of depression also does not prove a separate anti-insomnia action.
An insomnia assessment separates subjective distress from measured sleep and records daytime function, circadian timing, medication, substance use, and comorbidity. Relaxation during a session, change in a single questionnaire item, or an electroencephalographic band shift is not durable remission. Stronger trials preregister insomnia as a primary outcome, use credible sham and test masking, and report sleep diaries, actigraphy or polysomnography, follow-up, and adverse events. Afternoon or evening delivery time and sleep loss before treatment can themselves influence excitability and seizure risk, so they are protocol variables rather than background noise.
rTMS does not splint an upper airway or correct nocturnal hypoxemia and cannot replace diagnosis and airway-directed treatment of obstructive sleep apnea. Multicomponent cognitive behavioral therapy for insomnia remains a core evidence-based treatment for chronic insomnia. Any use of rTMS for sleep complaints must be judged against the local regulatory status, exact product, and professional assessment. Markedly reduced sleep accompanied by unusual energy, impulsivity, or elevated mood during a course should prompt assessment for a manic-spectrum change, not be celebrated as a restored need for less sleep.