Transcranial magnetic stimulation (TMS) is a non-invasive method of treating depression that uses repetitively pulsed magnetic fields to generate an electrical current in a targeted region of the brain.   

The magnet energy extends only 2 to 3 cm into the brain in an area of about size of 50 cent piece. It penetrates farther, 4-6cm with dTMS (deep TMS), using a Brainsway machine.

Unlike medication, this is a non-systemic treatment that has virtually no side effects, yet often has long term benefit. Transcranial magnetic stimulationRunning electricity through a copper coil creates a magnet.  A static magnet placed on the brain does nothing.   However, if a powerful magnet goes off and on rapidly, electrical changes occur in the neurons within the magnetic field.   The neurons are either excited or inhibited, depending on the position of the coil, the position of the neurons and the type of pulses administered.  Each pulse is 280 microseconds long in a Magpro Machine by Magventure, and pulse width is similar in other brands. Where and how to treat was basically discovered empirically.  There is one reasonable theory of why TMS works, though it is probably more complicated.   The left brain is hypoactive during depression.   Excitatory TMS treatment of the left DLPFC (dorsolateral prefrontal cortex) causes an increase in blood flow, metabolism and activity for about 45 minutes. Then activity returns to normal and there is no detectable difference in the body.    After many treatments, usually 5 to 15, the depression begins to lift.  Perhaps we are inducing he left brain to come to normal activity.  . It is harder to explain why inhibitory treatment on the right works for depression as well.  Maybe it inhibits that area until both sides of the brain are matched in activity?  Also, there are individuals who respond best to left inhibitory treatment and others who respond to right excitatory.


  • I usually do at least a 1 hour evaluation—including history, blood pressure and weight. If the patient is eager to start, and/or they have come a long ways for the appointment,   I often combine evaluation, obtaining the MT (motor threshold) and first treatment all in one visit.  But I don’t charge for all that.
The first treatment visit involves:
  • Transcranial Magnetic Stimulation - focal field pattern

    By MistyHora at the English language Wikipedia, CC BY-SA 3.0,

    Finding the MT (motor threshold), which takes about ½ hour. This is done only on the first visit, and is repeated periodically.  The MT is the amount of stimulus (magnetic energy, technically Teslas) applied to the hand area of the motor strip sufficient to cause motor movement in the thumb or fingers of the hand contralateral to the stimulus.  We use the MT to determine how much stimulus to use for the treatment (anywhere from 80% to 120% of MT, depending on protocol).
  • Each patient has his own cap, which is positioned the same on each visit. After MT is determined and other measurements are made, the place of treatment is determined and marked on the capsule.   Thereafter we can easily repeat the treatment in the same place each time.
    • The place of treatment is the DLPFC (dorsolateral prefrontal cortex), usually determined by going 6cm anterior in the sagittal plane to the “hotspot”—the area of the motor cortex where we elicited movement in the hand. We also determine it by finding F3 and F4 by the Beam method.   F3 and F4 are roughly equivalent to the left and right DLPFC. They are positions of placement for an EEG on the 10-20 system.   If these two positions are very far apart we may elect to treat between them.  Research indicates that the Beam method is probably more accurate than the “6cm rule.”
  • The treatment itself lasts about 30 minutes. If the patient is unipolar (or only mildly polar), the treatment is usually HF-L
[high frequency (excitatory) treatment on the left DLPFC (dorsolateral prefrontal cortex)—the left forehead].   A train of 40 pulses is administered at 10 to 20 pulses per second.   There is a 10 to 20 second interval between trains, primarily so the copper coils do not overheat. . —iTBS (intermittent theta burst stimulation) is a new form of  excitatory TMS which only takes 3 minutes and 8 seconds.   We usually reserve it for patients who cannot stay for a longer treatment, or if we are doing multiple treatments (in different places) in the same session.   I think it raises the risk of a seizure a little.  Alvaro Pascual-Leone, MD routinely demonstrated iTBS on students during his weeklong course at Harvard.  However, during a course in 2012, a student (a neurologist, jet lagged coming in from Europe) had a seizure during the demonstration.  Alvaro has stopped demonstrating iTBS.    It is possible that this was just one of the rare seizures associated with TMS.  Since iTBS has not yet been used extensively, it could be that iTBS is more likely to induce a seizure.   This is quite possible since iTBS is as effective (or more so) in 3 minutes as the usual 30 minute protocol. —CTBS (continuous theta burst stimulation) is inhibitory, and treatment only takes 40 seconds. [The Neuronetics machine cannot do TBS, but Brainsway, Magventure and Magstim machines can.]
  • Treatments end up taking about 45 minutes, when you add in getting in the chair, setting up, etc. You can count on one hour out of your day, plus travel time.  Since this can be inconvenient, so we are using TBS (theta burst stimulation) more and more.
  • If the patient is very bipolar, we usually do a low intensity (inhibitory) treatment on the right DLPFC (dorsolateral prefrontal cortex).
  • TMS is done 5 days per week–at any time of day that is mutually convenient. But 4 days/week would probably also work.  There are no studies comparing 30 treatments at 3 days per week vs 30 treatments at 5 days per week.  Alvaro Pascual-Leone, MD at Harvard says that he originally intended to treat 7 days per week almost 20 years ago in Valencia, Spain.  The main doors to the building were locked on weekends, and the only access to the TMS lab was through the morgue.   The ethics committee flatly ruled that patients with depression should not walk through a morgue twice a week.  He says that is how treatments 5 days per week came about.
Sometimes we treat on weekends, especially in the Mt Kisco office–if we can prevail upon a technician to work the weekend.  Personally, I think 7 days a week is better—but on little evidence.  People with TRD (treatment resistant depression) feel a bit hopeless, so obtaining success quickly is important.  If an individual takes 18 treatments to begin to demonstrate obvious benefit, we get there a lot more quickly with 7 days per week treatment than 5 days per week—2½ weeks instead of 3½ weeks.       —It seems that the total number of treatments is what is necessary, not the frequency.  If a particular person were to need 50 treatments to reach complete euthymia, it would take 10 weeks at 5 times per week, or 20 weeks if 2 times per week.   A treatment course is usually approximately 30 treatments [although this varies considerably].       A large Japanese group, because of long commutes, only treats 3 days per week.  They have the same good results as we do, but it takes about 10 weeks instead of 6 weeks. Two patients who I had treated for years [a 40 y.o. woman with almost incapacitating unipolar depression, and a 29 y.o. woman with Bipolar II Disorder] achieved remission after 23 treatments.   This is the first time either of them had experienced a completely normal mood.  The 23 treatments were over 11-12 months, however!  Because of logistics they only had about 2 treatments per month. —-We prefer to do the treatments between 9AM and 6PM.  We try, however, to make accommodations for people who have to work, and sometimes do treatments at 7:30 AM or 7PM.
  • Doing treatments two or more times per day SEEMS to accelerate response, but this is far from proven. When response time is critical, the patient is extremely suicidal, or the commute is very long, we often do treatments twice or more per day.
COMMON SIDE EFFECTS:   Side effects are usually mild and transitory.   Typically the patient has a treatment and then goes to work as usual.   In the early treatments we start low and go up slowly, so if it becomes uncomfortable we can lower the intensity and/or move the coil a little.    [With the Neurostar machine, because their computer program required one to start at 80% of MT, it was difficult to start at a very low intensity.  Now we can start as low as we want.] Common side effects can include:
  • Headache—the most common side effectit usually does not occur after 2-3 treatments, as the scalp muscles accommodate. We often give Celebrex 200mg before the first couple of treatments.
  • Scalp discomfort at the site of stimulation. Where the coil is in contact with the head the magnetic field is strongest.   Like light, magnetic energy exponentially decreases with distance—1/distance squared.
  • Tingling, spasms or twitching of facial muscles—if very uncomfortable, we decrease the intensity and/or move the coil a centimeter.
  • Lightheadedness – infrequent and transient
  • Discomfort from noise during treatment— Use ear plugs, which we provide.
  • We have had two patients [out of 70+] who became intensely tired for one to two days after each treatment. Both had high intensity TMS of left DLPFC.  One was an 18 y.o. with rapid cycling Bipolar II Disorder [years of depression and rapid cycling remitted with TMS.  MADRS 32 → 8.]. I had previously treated him for 2 years.  His mood disorder was amazingly brittle and treatment resistant.  The other was an 83 y.o. Bipolar II man [MADRS 25 → 1].   These patients were so successful after years of depression and unsuccessful medication, that in retrospect perhaps tiredness from TMS is a good sign.
  • Mild confusion, transient.
  • There seem to be no permanent side effects. Remember, the electrical activity on one side of the brain is altered for only 45 minutes after a treatment.  It then returns to baseline.
  • Mania, particularly in people with bipolar disorder. This risk can be largely avoided by doing inhibitory (“low intensity”) TMS on the right DLPFC.
  • Hearing loss due to inadequate ear protection during treatment–so we insist people wear ear plugs. The pulses are very brief, so they do not sound as loud as they actually are.  There is only one report of someone having a decrement in hearing secondary to TMS.  Because he already had a significant hearing problem, he assiduously used ear plugs, but one day an ear plug fell out without him realizing it.
  • Seizures—rare. The incidence seems to be about 1 seizure out of every 10,000 treatments.
    • Some of these occurred at high intensities, before current safety guidelines.
    • Apparently there was only one seizure out of the first 300,000 treatments done with the Neurostar machine. It occurred because the doctor accidentally treated the patient on the motor cortex instead of on the DLPFC (dorsolateral prefrontal cortex).
    • Some were apparently fainting episodes or some other event—not real seizures.
    • One seizure occurred 4-5 hours after a TMS treatment–so not related. He had a seizure disorder which was not under complete control.  He had a seizure approximately every 2 weeks and it had been about 2 weeks since his last seizure.  Even the author of the report said that the seizure was probably not secondary to TMS.
    • I read a report recently of an event wherein the doctor arrived in time to observe it (often only the technician observes the events called seizures).  For many reasons, it was not a seizure, including that the patient would squeeze his hand on command during the event, he had no confusion or headache after the event, and he remembered what was said during the event.
    • If someone has a pre-existing seizure disorder, a course of TMS will precipitate a seizure in 1-2%. Inhibitory TMS is used to treat epilepsy—doing the treatment over the epileptic focus, if it can be found.  Someone with a seizure disorder can safely be treated with inhibitory (“low intensity”) right DLPFC treatmentwhich actually seems to work as well as excitatory left DLPFC treatment for depression.   Since the initial papers used left DLPFC there has been little research using other targets.  There are at least 4 other targets that would theoretically be better.  “New Targets for rTMS in Depression: A Review of Convergent Evidence”,  Jonathan Downar,  Jeff Daskalakis, Brain Stimulation 6 (2013) 231-240
CONTRAINDICATIONS: “Do you have any metal from the shoulders up?”
  • Ferromagnetic plates in the cranium (titanium is OK)
  • A clipped aneurysm – unless you know the clip was not ferromagnetic.
    • However, if the clip is in the opposite hemisphere, and the treatment is 1 Hz inhibitory, then it is probably quite safe.
  • There should be no ferromagnetic metal within 12 inches of the coil
  • Cardiac pacemaker:  Weigh risk vs benefit.   Magnets deactivate pacemakers.  The pacemaker is far enough away to be of no problem.  If the patient’s head is extended, then I measure the coil as being ≥ 12 inches from the pacemaker.  The main consideration would be to not accidentally allow the magnetic coil to get near the chest.  This is easy for us. The machine is in back of the partially supine patient.  We have difficulty extending the coil far enough to reach the forehead, much less the chest.  Magnetic pulses only occur during active treatment—when the coil is rigidly fixed.  We could also do TBS, wherein the stimulus intensity can be 80% of MT instead of 120%.
  • Seizure disorder – this is safe if one uses low intensity treatment on the right
  • Metal dental braces could heat up—see if patient feels the heat. A longer inter-train interval  and/or fewer pulses per seconds would obviate this problem.  Dental fillings are OK.
  • Facial tattoos—be careful. If done with ferromagnetic material, the tattoo could heat up
  • A tattooed eyebrow could have ferromagnetic material and it is close to the coil.
  • Permanent ferromagnetic sutures from plastic surgery
  • Usually one sees some response in the 2nd week.
  • If there is no benefit after 3 weeks (15 treatments), then only 1/5 will have a significant response at the end of 30 treatments.
  • If there is no response in the 3rd week (10-15 treatments), my preference is to #1: stop treatment and save the patient money and time or #2: change the parameters—e.g., treat the other side. Obviously, I prefer #2, as I am convinced many people would be responders if we had the luxury to do 60 to 100 treatments and try many parameters.
  • When having good benefit in the first weeks of treatment, it is not uncommon for the patient to have a severe relapse over a weekend. We warn in advance so they do not become too disheartened by a setback.   Usually they are out of it after the 2nd or 3rd treatment on Tuesday or Wednesday.
  • There are reports of delayed response. Infrequently,  only a few days or weeks after  a series of TMS treatments do some patients experience a significant and long lasting improI had a patient to whom I gave over 50 treatments over a year without any benefit.  But a few months later she became, at 60 y.o., euthymic for the first time in her life.
  • When I took the TMS course at Duke in 2011, I was astonished to hear Holly Lisanby (then chairman of psychiatry at Duke) say that “the benefit is permanent”. Having already treated over a dozen people, within moments I knew what she meant, though I had never thought of it that way.
  • When there is a relapse the patient does not need another 30 treatments. Often 2 or 3 will suffice.   A permanent or semi-permanent change has been made in the brain.
  • Since depression is a chronic illness there are likely to be relapses. A few of our patients were never euthymic in their life prior to TMS, yet remained euthymic over 3 years after their last treatment.  [They are on medication, which is important to prevent relapse.]  We expect relapse eventually, however.   Some people relapse more regularly and need touch up treatments.  Some patients need maintenance treatment—e.g., 1 to 3 treatments once a month.  Only infrequently do we have a patient who we maintain with once a month prophylactic treatment.