Melodic intonation therapy

1. Introduction

Version followed: the three-level hierarchical MIT programme (Elementary, Intermediate, Advanced) of the American MIT manual (Helm-Estabrooks, Nicholas & Morgan, 1989), as reported in secondary sources that cite it; the manual itself is not drawn on here [uncertain on exact wording, item lists and scoring details]. Variants (Schlaug/Norton intensive protocol, Dutch RCT protocol, French TMR, palliative MIT) are named where they differ.

Melodic intonation therapy (MIT) is a speech and language therapy (SLT) method for adults with non-fluent (Broca's-type) aphasia after left-hemisphere stroke. It was introduced by Martin Albert, Robert Sparks and Nancy Helm (later Helm-Estabrooks) in the early 1970s (Albert, Sparks & Helm, 1973; Sparks, Helm & Albert, 1974, both cited in Zumbansen et al., 2014; the Boston VA setting often attributed to them is [uncertain]). Everyday phrases are sung slowly on two pitches that follow the phrase's natural stress, while the left hand taps once per syllable; therapist support is then withdrawn step by step, the melody is replaced by speech-song (Sprechgesang) and finally by normal speech [source: Zumbansen et al., 2014; ASHA Leader, 2012; Merrett et al., 2014].

Core idea of change: the developers' account is that intoning and left-hand tapping engage preserved, largely right-hemisphere, music and sensorimotor networks, giving access to speech output that the damaged left hemisphere no longer provides; Albert described the aim as helping people reach the brain's non-standard speech output systems [source: ASHA Leader, 2012]. Schlaug and colleagues add that slow, sustained syllables, prosodic chunking, and tapping as a rhythmic pacemaker may support auditory-motor coupling, while stating that the right-hemisphere account is not proven [source: Schlaug, Marchina & Norton, 2008; Merrett et al., 2014].

Professional frame and adjunct status:

MIT is a rehabilitation technique delivered within medical/rehabilitation care, normally by a qualified speech and language therapist (in trials, experienced SLTs trained in the protocol) [source: Van der Meulen et al., 2016]. In the UK that means an HCPC-registered SLT (RCSLT is the professional body); in the US an ASHA-certified SLP (CCC-SLP) [synthesis: standard registration routes, not from an MIT source] [uncertain]. Music therapists with Neurologic Music Therapy (NMT) training also use an MIT-style technique within the NMT taxonomy; this listing was only confirmed in an exam-prep study guide [uncertain]. NMT training does not let non-music-therapists practise beyond their own licence [source: ArtEZ NMT programme page].
A psychotherapist should not deliver MIT as a stand-alone treatment of aphasia. Your role is to refer to SLT, work alongside the SLT, support mood, adjustment and motivation, and, with the SLT's agreement, prompt the patient's own trained phrases in your sessions [synthesis].
Evidence status as sources give it: the Cochrane review of SLT for post-stroke aphasia (Brady et al., 2016) found SLT improves functional communication versus no SLT and that higher intensity/dose/longer duration may help, though with more drop-outs; the review does not report on MIT specifically. A Dutch pilot RCT in chronic aphasia found gains on trained items only, no between-group generalisation, and loss at follow-up; the authors report more favourable results in their earlier subacute RCT [source: Brady et al., 2016; Van der Meulen et al., 2016]. Zumbansen et al. (2014) found only five studies that followed the original protocol and rationale.
Hearing safety: MIT is sung at speaking-to-moderate volume, one to one, without amplification; keep any accompanying keyboard soft (one telepractice study kept a piano at low volume) and check hearing aids are in and working before sessions [source: Darland, 2021; synthesis].
Musical trauma triggers: the manual tradition avoids familiar song melodies anyway (two-note contours only) [source: Darland, 2021, citing the MIT authors]; still ask about songs, hymns or musical settings tied to grief, bereavement or trauma before using any melody or accompaniment (e.g. palliative variants with fixed tunes), and stop if singing itself provokes distress [synthesis].

2. Framework

MIT is a single structured method with a candidacy step, a three-level linguistic hierarchy and a scoring system that decides advancement [source: Darland, 2021, reporting the Helm-Estabrooks manual; Pro-Ed sample page; ASHA Leader, 2012].

Phase 0, candidacy and setup: confirm the patient fits the profile (non-fluent output, relatively good auditory comprehension, poor repetition, left-hemisphere stroke, adequate attention and emotional stability), choose and melodically code target phrases, agree schedule and home practice [source: Pro-Ed sample page; Van der Meulen et al., 2016].
Level 1, Elementary: short, high-probability words and phrases (2–4 syllables) sung with the therapist and left-hand tapping; five steps from humming to answering a question with the sung phrase. Goal: produce the intoned target with progressively less support [source: Darland, 2021; ASHA Leader, 2012].
Level 2, Intermediate: longer phrases; unison with fading, delayed repetition, then answering a related question without tapping. Goal: hold the phrase across a delay and retrieve it to a question [source: Darland, 2021].
Level 3, Advanced: longer or phonologically more complex phrases; moves from intoning to Sprechgesang (exaggerated speech prosody), to normal speech, then to answering questions. Goal: transfer to spoken, propositional use [source: Darland, 2021; Pro-Ed sample page; ASHA Leader, 2012]. The French TMR variant has no guided return to normal speech [source: Zumbansen et al., 2014].
Format: individual, face to face, therapist seated beside or opposite the patient to hold and tap the left hand and give a visual model; picture cards for targets [source: Schlaug et al., 2008; Pro-Ed sample page; seating detail is synthesis].
Dose: the manual tradition says as often as possible [source: Darland, 2021]; the original description gave frequent sessions over 3–6 weeks (Zumbansen's paraphrase reads at least 3 a week [uncertain]); Schlaug's intensive protocol used 1.5 h/day, 5 days/week, with patients assessed at 40 and 75 sessions [source: Schlaug et al., 2008]; the Dutch RCT planned 5 h/week for 6 weeks (minimum 3 h face to face, the rest app-based home practice) [source: Van der Meulen et al., 2016]. Reported schedules vary from weeks to many months [source: Zumbansen et al., 2014].
Criteria for moving between levels: advance when the patient reaches about 90% of possible points; Darland reports this both as "across several sessions" and as an average over ten consecutive sessions, citing secondary texts and Norton et al. (2009) [uncertain: confirm against the manual]. Clinician judgement of readiness is also used [source: Darland, 2021]. Within a level, a phrase that fails after the permitted retry or cue is dropped and a new phrase is started from the first step [source: Darland, 2021].

3. Guiding principles

Select candidates strictly: MIT is for non-fluent aphasia with relatively preserved comprehension; misapplying it to other profiles is a common reason it fails. Practical meaning: screen before starting and refer elsewhere if the profile does not fit [source: ASHA Leader, 2012; Pro-Ed sample page].
Severity and chronicity are not exclusions: Albert said "severity is not a limiting factor". Practical meaning: do not rule out someone years post-stroke or with a severe stereotypy if other criteria fit [source: ASHA Leader, 2012].
Intone, do not sing tunes: two fixed pitches (about a minor third or fourth apart), higher on naturally stressed syllables, no familiar melodies. Practical meaning: code every phrase from its spoken stress pattern [source: Zumbansen et al., 2014; Darland, 2021].
Slow and legato: roughly one syllable per second, continuous voicing, not staccato. Practical meaning: stretch vowels and connect syllables; choppy singing is a known error [source: Schlaug et al., 2008; ASHA Leader, 2012].
Tap the left hand on every syllable: tapping is treated as a key component, and omitting it is a known error in the original method. Practical meaning: one tap per syllable on every intoned item at the levels that use it [source: ASHA Leader, 2012; Merrett et al., 2014]. Some variants drop tapping [source: Zumbansen et al., 2014].
Fade support systematically: unison, then fading, then repetition, then delay, then question-answer. Practical meaning: never remove two supports at once; back up one step on failure [source: Darland, 2021].
Copious, varied, functional material: avoid rote memorising a few items; use everyday, personally relevant phrases chosen with patient and family. Practical meaning: build a large phrase bank and rotate it [source: Zumbansen et al., 2014; Van der Meulen et al., 2016].
Return to speech: the original aim is normal speech via Sprechgesang, not permanent singing. Practical meaning: plan the transition from the start [source: Zumbansen et al., 2014].
Intensity matters: frequent, intensive practice, with home practice to raise dose. Practical meaning: agree a realistic schedule the patient can sustain [source: Van der Meulen et al., 2016; Brady et al., 2016].
Measure every response: step-by-step scoring drives decisions. Practical meaning: score each attempt in session, not from memory afterwards [source: Darland, 2021].
Adjunct, not replacement: MIT sits inside stroke rehabilitation and SLT care [synthesis].

4. Intervention inventory

All example language is illustrative unless quoted. Step structure and scoring follow Darland (2021), who reports the Helm-Estabrooks et al. manual; point values and wording should be checked against the manual before clinical use [uncertain]. Items 4.1–4.5 are setup components used throughout; 4.6–4.8 are the levels; 4.9–4.14 are monitoring, home practice, variants and the psychotherapist's adjunct role.

4.1 Candidacy screening

Purpose: identify who is likely to respond; done before any treatment.
Steps: (1) Confirm stroke aetiology and a left-hemisphere lesion with an intact right hemisphere (Naeser & Helm-Estabrooks, 1985, profile: lesion involving or undercutting Broca's area, no large Wernicke's-area lesion) [source: Pro-Ed sample page]. (2) Check output is non-fluent or reduced to a stereotypy that is at least partly suppressed when singing. (3) Check auditory comprehension is relatively good. (4) Check repetition is poor, even for single words, and articulatory agility is poor (Helm, 1978, characteristics; BDAE rating of 3 or less) [source: Pro-Ed sample page]. (5) Check emotional stability, motivation and attention span are adequate, and that the person can commit to an intensive programme [source: Pro-Ed sample page; Darland, 2021]. (6) Operational cut-offs used in research: the Dutch RCT required under 50 words/min, AAT repetition and articulation limits and minimum comprehension scores; Darland used WAB-R fluency, comprehension and repetition cut-offs [source: Van der Meulen et al., 2016; Darland, 2021]. Exclusions in the Dutch RCT included bilateral lesions, previous intensive MIT, severe hearing loss and relevant psychiatric history.
Example language (illustrative): "I'm going to sing a short line and ask you to hum along — it helps me see whether this approach suits you."
Watch for: fluent aphasia, preserved repetition, poor comprehension, bilateral damage: refer for a different approach. Severity alone is not a reason to exclude [source: ASHA Leader, 2012].

4.2 Choosing and coding target phrases

Purpose: build the stimulus bank for each level.
Steps: (1) With patient and family, list high-frequency, functional, personally relevant phrases (greetings, needs, names, hobbies) [source: Darland, 2021; Van der Meulen et al., 2016]. (2) Grade by length: Elementary 2–4 syllables (one or two words), avoiding consonant clusters and liquids at first; lengthen and add phonological complexity at later levels [source: Darland, 2021; Schlaug et al., 2008]. (3) Say each phrase naturally and mark stressed syllables. (4) Assign two pitches about a minor third or perfect fourth apart: stressed syllables high, unstressed low; set rhythm to quarter and eighth notes, louder on stressed syllables [source: Zumbansen et al., 2014; Darland, 2021]. (5) Prepare a picture card for each item [source: Pro-Ed sample page; Schlaug et al., 2008]. (6) Keep the bank large and varied, and reserve some untrained phrases for probing generalisation [source: Zumbansen et al., 2014; Van der Meulen et al., 2016]. In the Dutch RCT at least half of therapy time used a standard protocol list, the rest personal phrases [source: Van der Meulen et al., 2016].
Example language (illustrative): "Which things do you most want to be able to say at home? Let's pick ten."
Watch for: familiar song tunes creeping in (avoid) [source: Darland, 2021]; phrases with emotional charge that upset the patient [synthesis].

4.3 Intoning: pitch, rate, continuous voicing

Purpose: the core vocal technique for every intoned item.
Steps: (1) Choose a comfortable low note for the patient as the lower pitch [source: Darland, 2021]. (2) Sing each syllable at about one per second [source: Schlaug et al., 2008; Merrett et al., 2014]. (3) Keep voicing continuous and legato, lengthening vowels; not staccato [source: ASHA Leader, 2012]. (4) Keep the contour fixed for that phrase across sessions.
Example language (illustrative): therapist sings "COF-fee" on high–low, slowly, voice flowing between syllables.
Watch for: rushing, choppy delivery, drifting pitch; pitch accuracy by the patient is not required (one study kept the keyboard quiet so patients were not drawn to pitch) [source: Darland, 2021].

4.4 Left-hand tapping

Purpose: rhythmic pacing and, on the developers' account, right-hemisphere sensorimotor priming [source: Schlaug et al., 2008].
Steps: (1) Ask permission to hold the patient's left hand [synthesis: consent]. (2) Hold it and tap it once per syllable as you intone, in time with the syllables [source: Pro-Ed sample page; ASHA Leader, 2012]. (3) At Elementary level the therapist taps the patient's hand; in the first steps of Intermediate and Advanced levels the patient taps; tapping stops at the question steps and normal-speech step [source: Darland, 2021]. (4) In telepractice the patient taps their own left hand [source: Darland, 2021].
Example language (illustrative): "May I hold your left hand? I'll tap it with each sound we sing."
Watch for: tapping out of sync with syllables; hand pain, oedema, fragile skin or bruising; withdrawal from touch (stop and use self-tapping) [synthesis].

4.5 Visual and phonemic cueing

Purpose: support imitation within steps.
Steps: (1) Sit so the patient can see your face and mouth; the therapist gives an auditory and visual model to synchronise with [source: Merrett et al., 2014]. (2) Present the picture card with the phrase [source: Pro-Ed sample page]. (3) Where the step allows a cue, give the first sound of the first word [source: Darland, 2021].
Example language (illustrative): therapist mouths and intones the initial sound only: "Wa…"
Watch for: more cues than the step permits (score down or back up a step) [source: Darland, 2021].

4.6 Elementary level (Level 1), five steps

Purpose: first production of short intoned phrases with maximal support; used for the first phrases and for patients with minimal output.
Steps [source: Darland, 2021; NHS West Suffolk leaflet, 2019]: (1) Humming: show the picture, hum the phrase's melody while tapping the patient's left hand; invite them to hum along. (2) Unison intoning: intone the words on the same melody with tapping, repeating until the patient sings with you; scored (1 point reported). (3) Unison with fading: start together, then drop your voice out partway while the patient continues alone; move on when they complete it. (4) Immediate repetition: intone the phrase alone, the patient then repeats it alone with tapping; one phonemic cue allowed (2 points without cue, 1 with; discontinue the phrase if it fails with a cue). (5) Response to a probe question: intone a question whose answer is the target phrase, the patient answers in intoned form; scored like step 4.
Example language (illustrative): hum "mm-MM" while tapping; then intone "WA-ter" together; fade on "-ter"; then: "Now you"; then intone "What do you want?" with the answer expected as "WA-ter".
Watch for: the patient joining from humming to words; a stereotypy replacing the target (return to unison); success in unison but collapse on fading (stay at step 3 for more trials).

4.7 Intermediate level (Level 2), three steps

Purpose: build retention across a delay and retrieval to a related question; for patients who meet the Elementary criterion.
Steps [source: Darland, 2021]: (1) Unison with fading: intone together with the patient tapping, then fade out; scored. (2) Delayed repetition: you intone the phrase, wait several seconds (an exam-prep guide gives 6 seconds [uncertain]), then signal the patient to produce it alone, typically twice with a short gap; one retry allowed; discontinue the phrase if the retry fails. (3) Response to a related question, without tapping: ask a question the phrase answers indirectly (e.g. target about wanting a drink, question about what kind); if it fails, go back to delayed repetition and try again; discontinue if that fails.
Example language (illustrative): intone "I want a DRINK", pause, nod: "Your turn"; then speak: "What would you like?"
Watch for: success only immediately after the model (delay too long: shorten, then rebuild); answering with a single word (acceptable at this level only if it fits the scoring the manual sets [uncertain]).

4.8 Advanced level (Level 3), four steps including Sprechgesang

Purpose: transfer from intoning to normal prosody and spoken, propositional use.
Steps [source: Darland, 2021; Pro-Ed sample page; ASHA Leader, 2012]: (1) Sprechgesang in unison: intone the phrase, pause, then present it in Sprechgesang (rhythm and stress exaggerated, no fixed pitch, "half-way between singing and speech" [source: Zumbansen et al., 2014]); patient joins in unison with tapping; retry allowed. (2) Sprechgesang with delayed repetition: present in Sprechgesang, pause, patient repeats alone. (3) Spoken repetition, no tapping: say the phrase with normal prosody, pause briefly, patient repeats. (4) Questions: ask specific questions about the target (back up to step 3 if needed), then less specific and general ones, with no backup on general questions.
Example language (illustrative): Sprechgesang "I'm GO-ing HOME to-DAY" with stressed syllables stretched; then normal: "I'm going home today"; then: "What's happening this afternoon?"
Watch for: falling back into singing when spoken repetition gets hard (return to step 2); spontaneous use of the phrase outside the task (good sign) [synthesis].

4.9 Scoring and session records

Purpose: objective decisions on advancing, backing up or dropping a phrase.
Steps: (1) Score every attempt at each step as it happens, using the manual's point values. (2) Total points earned versus possible per session. (3) Track per-phrase failures; drop a phrase after the allowed retry/cue fails and start a new one at step 1 [source: Darland, 2021]. (4) Probe trained and untrained phrases periodically (e.g. pre-treatment probes, end of treatment, a maintenance probe weeks later) [source: Darland, 2021; Van der Meulen et al., 2016].
Example language (illustrative): none needed; keep scoring unobtrusive.
Watch for: rising trained-item scores with flat untrained probes (limited generalisation, as in the Dutch RCT) [source: Van der Meulen et al., 2016].

4.10 Level advancement decision

Purpose: move to the next level at the right time.
Steps: (1) Average session percentage scores; (2) advance at about 90% sustained across consecutive sessions (reported as several, or ten, sessions) [source: Darland, 2021; uncertain]; (3) add clinician judgement of readiness [source: Darland, 2021].
Watch for: advancing on a single good day; frustration from staying too long at a mastered level [synthesis].

4.11 Home practice

Purpose: raise dose between sessions.
Steps: (1) Provide home practice materials for trained phrases (Schlaug's patients had them) [source: Schlaug et al., 2008]. (2) In the Dutch RCT, when face-to-face time was short, patients practised with a tablet/iPod app 2–7 h/week and patients or relatives logged practice time [source: Van der Meulen et al., 2016]. (3) Train carers to model and tap gently, and not to drill to exhaustion [synthesis].
Example language (illustrative): "Practise these five with the app twice a day, and note the minutes on this sheet."
Watch for: carers pushing beyond fatigue; practice done in a rushed or staccato way [synthesis; ASHA Leader, 2012].

4.12 Intensive Schlaug/Norton protocol (variant)

Purpose: high-dose delivery used in Boston imaging studies.
Steps: one-to-one sessions of 1.5 h/day, 5 days/week, until all three levels are mastered; patient taps the left hand per syllable; 1 syllable/s; two pitches with stressed syllables higher; continuous voicing; unison and antiphonal (alternating) repetition; continuous feedback; picture cues; phrases from 2 syllables up to longer requests [source: Schlaug et al., 2008].
Watch for: fatigue and drop-out with high-intensity schedules [source: Brady et al., 2016].

4.13 Other published variants

French TMR (Van Eeckhout & Bhatt): intoned speech taught as an occasional aid when the patient gets stuck; artificial stress emphasising function words; tapping with any body part; no guided return to normal speech; long, less intensive period [source: Zumbansen et al., 2014].
Palliative MIT: trains a few ready-made sentences for basic communication; Baker (2000) gave each sentence its own melody and harmonic accompaniment as a memory cue, without tapping; Hough (2010) also omitted tapping [source: Zumbansen et al., 2014; Merrett et al., 2014].
Telepractice adaptation: patient self-taps, therapist plays a quiet keyboard off-camera [source: Darland, 2021].
Multimodal MIT plus group socialisation (single case) [source: AJSLP 2018 case study abstract].
Purpose and watch for: name the variant in your notes; do not mix TMR's "aid when stuck" aim with the original's return-to-speech aim without saying so [synthesis].

4.14 Psychotherapist's adjunct role [synthesis]

Purpose: support a client in MIT without delivering MIT.
Steps: (1) With consent, liaise with the treating SLT: current level, trained phrases, whether to prompt them. (2) In sessions, allow time; use yes/no questions, pictures and written key words for your own communication. (3) If agreed, invite the client to use a trained phrase in its current form (intoned or Sprechgesang). (4) Address post-stroke low mood, grief, shame about singing, and carer strain; screen for depression with aphasia-friendly tools and refer as needed.
Example language (illustrative): "Would it help to sing it the way you practise with your speech therapist?"
Watch for: you becoming a second, unsupervised MIT trainer; contradicting the SLT's hierarchy.

5. Handling problems

Sources address few problems directly; items are marked.

Wrong candidate (fluent aphasia, poor comprehension, preserved repetition): the most common misuse. Response: stop MIT and refer for a better-matched SLT approach [source: ASHA Leader, 2012].
Delivery drift (staccato singing, skipped tapping): known reasons the method fails. Response: review a recording against the coding sheet; restore legato voicing and one tap per syllable [source: ASHA Leader, 2012].
Stuck at a step: back up one step, as the hierarchy allows (e.g. from related question to delayed repetition); if the retry fails, drop the phrase and start a new one at step 1 [source: Darland, 2021]. Shorten or simplify phrases (fewer syllables, no clusters) [source: Darland, 2021; synthesis on applying it].
Stereotypy or perseveration intruding: return to unison and humming, slow down, change to a phrase with different sounds [synthesis; the Pro-Ed sample notes stereotypies are at least partly suppressed when singing].
Gains limited to trained items or fading after therapy ends: expected in some patients; Dutch chronic RCT showed trained-item gains not maintained. Response: widen and rotate the phrase bank, probe untrained items, plan maintenance practice [source: Van der Meulen et al., 2016; maintenance plan is synthesis].
Fatigue, non-adherence and drop-out at high intensity: Cochrane notes higher drop-outs with intensive therapy. Response: shorter, more frequent sessions, app-based home practice, scheduled rests, involve carers [source: Brady et al., 2016; Van der Meulen et al., 2016; scheduling is synthesis].
Embarrassment or resistance about singing ("childish"): explain the rationale briefly, show the planned transition to speech, start with humming in private, use adult, personally meaningful phrases [synthesis].
Overwhelm, crying, catastrophic reactions or emotional lability: pause, lower demand to an easy successful step, end on success; distinguish post-stroke emotionalism from distress; screen for depression and refer [synthesis].
Rupture (patient feels drilled, corrected or rushed): name it, ask with yes/no options or a rating card, give the patient choice of phrases and pacing, repair before continuing [synthesis].
Touch problems (tapping): ask before holding the hand; offer self-tapping or tapping on a table if touch is unwanted or culturally uncomfortable; check the left hand for pain, injury, sensory loss or fragile skin (e.g. on anticoagulants); practise hand hygiene [synthesis; self-tapping is used in telepractice per Darland, 2021].
Hearing loss or hearing aids: confirm aids are working; face the patient; keep volume moderate; no amplified music [synthesis]. Severe hearing loss was an exclusion in the Dutch RCT [source: Van der Meulen et al., 2016].
Musical trauma triggers and grief: if a melody, hymn or accompaniment (in palliative variants) evokes bereavement or trauma memories, switch to neutral two-note intoning, stop and attend to the distress, and refer to the psychotherapist where appropriate [synthesis].
Clinician unsure of pitch: pitch accuracy matters less than a stable two-note contour; a quiet keyboard can set the notes [source: Darland, 2021; interpretation synthesis].
Telepractice limits: patient self-taps; use clear audio; carer may sit beside the patient [source: Darland, 2021; carer role synthesis].
Carers over-drilling at home: agree a fixed practice time and limit; teach the same fading steps the therapist uses [synthesis].
Medical change (new stroke symptoms, seizures, sudden worsening of speech): stop and seek medical review; MIT never replaces medical care [synthesis].

Sources

Sources:

Books and papers cited via the sources above (further reading, not drawn on) [uncertain]:

Helm-Estabrooks, N., Nicholas, M., & Morgan, A. (1989). Melodic Intonation Therapy (programme manual). Pro-Ed.
Albert, M. L., Sparks, R. W., & Helm, N. A. (1973). Melodic intonation therapy for aphasia. Archives of Neurology.
Sparks, R. W., Helm, N., & Albert, M. (1974). Aphasia rehabilitation resulting from melodic intonation therapy. Cortex.
Sparks, R. W., & Holland, A. L. (1976). Method: melodic intonation therapy for aphasia. Journal of Speech and Hearing Disorders.
Norton, A., Zipse, L., Marchina, S., & Schlaug, G. (2009). Melodic intonation therapy: shared insights on how it is done and why it might help. Annals of the New York Academy of Sciences.
Naeser, M. A., & Helm-Estabrooks, N. (1985); Helm (1978); Van der Meulen et al. (2014, subacute RCT); Baker (2000); Hough (2010).

← Previous · VIII-055Neurologic music therapy (Thaut)Next · VIII-057 →Orff music therapy