top of page

Music and the Brain: The Role of Musical Improvisation in Stroke Recovery

Updated: Jan 11, 2021

The Case of Trumpeter Louis Smith

By Hrayr Attarian, MD

Can jazz repair a damaged brain? This is neither a philosophical question nor an abstract statement; therefore, the article below is neither speculative nor theoretical. It is a concrete evaluation of the experimental data on music therapy, particularly jazz, in the treatment of brain injury due to a stroke. A stroke, also known as cerebrovascular accident (CVA), is akin to a heart attack as it is sudden and due to either a blockage to the blood flow or hemorrhage in a specific brain area. Rehabilitation that helps victims regain function includes complex and various modalities and, currently, music therapy is not a mainstay of this care. There is a small, yet robust, body of evidence in favor of using music therapy as an adjunct intervention in helping stroke victims recover function.

Louis Smith's first album as a leader, Blue Note, 1958

On August 20th of 2016 the great trumpeter and educator Louis Smith passed away from the sequelae of a stroke he suffered eleven years before. Smith, whom I met on November 16th, 2013, was 74 years old when he suffered a massive stroke that robbed him of his ability to both understand and produce any language, both verbal and written (a condition called aphasia), and paralyzed his right side. After weeks of intense, multi-modal therapy Smith demonstrated a modest degree of physical recovery without any improvement in his language skills. His high level of education, and strong family support, particularly that of his wife Lulu, together with his intellectually enriched living environment were all favorable odds for recovery. Him being in his eighth decade of life, together with the large size of the stroke, however, were major disadvantages. The rehabilitation team working with Smith told Lulu that there was nothing more that could be done and that Smith and his family should come to terms with his disabilities.

Seated L to R: Lulu and Louis Smith. Standing L to R: Dr. Gorden Krainen, Lynn Chenoweth, Dr. Hrayr Attarian.

In 2002, a pioneering researcher, Dr. Michael Thaut of Colorado State University in Fort Collins, used a metronome-like beat as an auditory cue during rehabilitation of patients with right arm weakness. Compared to those who did the same exercises without the metronome, those who were presented with the beat had significant improvement in both speed and accuracy of tasks performed by the affected hand. In 2007, in a larger study Taut and his colleagues showed greater improvement in gait after stroke when therapy was done to a rhythm versus without any sound prompts. Thaut also demonstrated that singing a series of words rather than repeating them without music improved their retention and reduced recall errors in patients with multiple sclerosis, another disease of the central nervous system.

Over in Europe, Finnish and English researchers led by Drs. Teppo Särkämö and David Soto, of the University of Helsinki and Imperial College London respectively, made equally intriguing discoveries. Stroke victims with visual loss showed greater recovery in their vision while they listened to their preferred music versus when they listened to music they did not like or nothing at all. As we have mentioned in previous issues of Music and the Brain, the scientists used a specialized medical imaging technique, called functional magnetic resonance imaging (fMRI). This showed an increase in activity in the area of the brain that is involved in language production as well. To determine musical exposure’s long-term benefits, Särkämö and Soto divided sixty patients with large strokes affecting similar areas of the brain into three random groups. One group listened to the music of their choice for a minimum of an hour per day for two months, the second group listened to books on tape for an hour or more per day for two months, and the third group was not offered any listening material. All sixty received the same standard of care rehabilitation otherwise. Using standardized testing, at one week, three months, and six months after the experiment the music group had significantly greater improvement in memory, mood, and attention, and significantly less confusion compared to the other two groups. All these studies used passive exposure to music and rhythm. What about active music playing? How does that affect stroke recovery? Can actually playing music, especially improvising, lead to greater improvement in brain function than just mere passive enjoyment of it?

University of Michigan speech and language pathologist Dr. Gordon Krainen, PhD, CCC-SLP and board certified music therapist Lynn Chenoweth, MT-BC refused to give up on Smith. They came up with a uniquely ingenious method to help Smith regain his language; they asked him to play trumpet. Demoralized by his disabilities, Smith was originally reluctant to perform music. Chenoweth coaxed him into it by asking him to teach her how to play the horn. Smith’s friends made a custom stand for Smith until he learned to play left handed and gained a bit more use of his right hand.

For jazz aficionados, Smith is a familiar name. Born Edward Louis Smith in 1931 in Memphis, Tennessee, he recorded four albums between 1957 and 1958 for the Blue Note label. Two were as a leader: Here Comes Louis Smith (Blue Note, 1958) and Smithville (Blue Note, 1958). Two others were as a sideman: guitarist Kenny Burrell’s Blue Lights (Blue Note, 1958) and pianist Horace Silver’s Live at Newport ’58 (Blue Note, 2007). He then took a twenty-year hiatus to devote himself to education. He taught at various public schools and universities in Georgia, Tennessee, Kansas, and Ann Arbor, Michigan—where he settled for the remainder of his life. He spent the last seventeen years of his educational career at Ann Arbor Junior High School and as a director of the University of Michigan’s Jazz Band. He also released a dozen superior records for SteepleChase Records.

Chenoweth worked with him on his trumpet playing, singing, and speaking using a technique called Melodic Intonation Therapy (MIT). MIT was designed by Boston VA hospital’s Dr. Martin L. Albert, Robert W. Sparks, and Nancy A. Helm. It is widely used by Dr. Michael Thaut of Colorado State University. MIT helped Smith retrieve some functional language, particularly language comprehension. Krainen, who is an accomplished pianist, helped Smith rehearse and helped him regain some speech by associating names of instruments with Smith’s favorite jazz musicians. Eight years later, Smith had progressed significantly in playing his trumpet left-handed and was beginning to improvise. It was then, at the American Speech Hearing and Language Association (ASHA) annual meeting on November 16th, 2013, that he, Chenoweth, Krainen and I presented his case. He demonstrated how he had mostly recovered his language comprehension and was even able to verbally express himself, albeit with considerable difficulty. Most importantly, he was no longer despondent and depressed, especially when able to pick up and blow his horn, although hesitantly, with his trademark lyricism.

The YouTube video below shows Smith playing in Chicago on November 16, 2013, at the annual ASHA convention after our seminar titled “The Restorative Power of Music for a Jazz Musician With Global Aphasia.” Dr. Gordon Krainen accompanies him on the piano.

Why and how would playing jazz help language recovery?

In 2016 Italian researchers in Milan and Padua studied active music therapy (aMT) that is based on free-improvisation in language impairment. They combined aMT with standard of care speech and language therapy (SLT) in a randomly chosen group of 10 stroke victims who suffered from aphasia. They compared them to a control group of 10 other patients with aphasia who only received SLT. All 20 received 30 SLT sessions over 15 weeks while the first group underwent an additional 30 sessions of aMT during those 15 weeks. The SLT sessions were 45 minutes in duration and aMT sessions were 30 minutes. During these the therapist and the patient played drums, xylophone, other percussion instruments, and also sang.

Spontaneous speech improved significantly in the group that received aMT while it did not in the group who only received SLT. In addition, half of the music therapy group showed significant improvement in certain quality of life measures while none of the other group did.

What was perhaps the most remarkable aspect of the study was that all participants had had their stroke 1–3 years prior to engagement in this therapy. So, theoretically they would have benefited significantly more had they been enrolled in music therapy shortly after their stroke. The brain recovers faster in the first few months after injury than a year or more later.

As we have discussed in previous “Music and the Brain” columns, making music, and particularly improvisation, activates areas of the brain that are closely related to language and speech. Furthermore, using rhythm instruments may organize and synchronize certain behaviors and movements—improving speech articulation. Lastly, the music therapist’s involvement in this improvisational process led to an empathetic relationship with the patient which enabled greater improvement in the latter’s speech.

It is quite fascinating that the compelling human need to produce music, although it serves no apparent evolutionary purpose, can mend the very organ responsible for its inception.


1) Koelsch, S. (2009). A Neuroscientific Perspective on Music Therapy. Annals of New York Academy of Sciences. 1169, 374–384.

2) Raglio, A., Oasi, O., Gianotti, M., Rossi, A., Goulene, K., Stramba-Badiale, M. (2016). Improvement of spontaneous language in stroke patients with chronic aphasia treated with music therapy: a randomized controlled trial. Int J Neurosci. 126:3, 235–242.

3) Särkämö, T., Soto, D. (2012). Music listening after stroke: beneficial effects and potential neural mechanisms. Annals of New York Academy of Sciences. 1252, 266–281.

4) Thaut, M.H., Kenyon, G.P., Hurt, C.P., McIntosh, G.C., Hoemberg, V. (2002). Kinematic optimization of spatiotemporal patterns in paretic arm training with stroke patients. Neuropsychologia. 40 (7), 1073–1081.

5) Thaut, M.H., Leins, A.K., Rice, R.R., et al. (2007). Rhythmic auditory stimulation improves gait more than NDT/Bobath training in near-ambulatory patients early poststroke: a single-blind, randomized trial. Neurorehabilitation and Neural Repair. 21(5), 455–9.

6) Thaut, M.H., Peterson, D.A., McIntosh, G.C. (2005). Temporal Entrainment of Cognitive Functions. Annals of New York Academy of Sciences. 1060, 243–54.

bottom of page