"Inaugural Musical Dynaformics Workshop" Event Report

"Inaugural Musical Dynaformics Workshop" was organized and held by Sony Computer Science Laboratory(Sony CSL) on October 29th 2023. This marked the world’s first international workshop on "Dynaformics" research for musicians.

Musical Dynaformics is a study by Tokyo Research Director and Senior Researcher Shinichi Furuya, who aims to realize the evolution and sustainability of musical expression by supporting musical expertise and musicians’ disorders.

The workshop was hosted by Dr. Furuya, and he brought together researchers worldwide who engage in research on music performance. There were research presentations and panel sessions by researchers at the forefront of their fields and shared their insights. Participants also enjoyed the lab tour and getting together party.

【Session 1 : Motivational factors driving learning in pianists】

Guest : Professor Maria Herrojo Ruiz (University of London, Goldsmith) UK

Learning biases and music performance anxiety

It remains yet to be understood how the changes in physiology in emotions and in cognitive processes contribute directly to the breakdown of performance of musicians, and what neural mechanisms underlie it . Professor Herrojo Ruiz has applied methods and techniques of "Computational Psychiatry" in her field of studies. It has contributed to better understanding of anxiety disorders.

She spoke about two aspects of her research; one is that anxiety is associated with negative learning biases, which means our tendency to learn faster from negative feedback and outcomes. The other is misestimation of uncertainty, which can be used to perhaps predict onset of a disease and the response to treatment.

Negative learning biases in musicians

According to the rhythmic-based formulations of predictive coding, anxious individuals are characterized by excessive beta oscillation which impairs updating their beliefs in line with predictive coding. It has been shown that individuals with anxiety learn, update and change their behavior faster when they receive negative feedback compared to positive feedback.

In an experiment, while there was no difference between learning from positive or negative feedback in the whole group of participants, anxiety scores in the individuals with very low music performance learned faster from negative feedback. The result is the opposite of what she had hypothesized. In another experiment, they used a representative piano performance task that required highly skilled pianists to learn from loss or reward about hidden target dynamics. As a result, highly anxious pianists learned faster from reward, while very low anxious individuals learned faster from punishment.

Pharmacological aids in music performance anxiety

Professor Ruiz has also introduced her fresh research results. 20 pianists participated in an experiment after taking propranolol, so called a beta-blocker. When pianists will get low reward on a trial and have to do maximum effort, they actually dropped effort in the beta blockers condition. As for the placebo condition, they produced a maximum effort much better. Here we can find an interesting interaction between effort and reward on the neural level, because the monkey studies never find any effect of reward in their tasks.

Professor Ruiz's team focused on the fact that theta waves are involved in many tasks, such as handling error and loss. Theta waves have been shown to be related to processing feedback in many tasks such as processing errors, processing losses etc. You may expect enhanced theta when you process losses or negative feedback. However, she found a reverse pattern that individuals who had beta blockers exhibited this abnormal response to high reward, whereas theta varied with losses or low reward or negative feedback. She will conduct a follow-up research to understand whether this effect regarding the reverse pattern of theta modulation during processing the reward feedback would explain their drop in performance levels with a low reward in the effort task.

【Session 2 : Exploring the Intersection of Music and Neuroscience: Insights from Singing Research】

Guest : Professor Boris Kleber (Aarhus University) Denmark

Professor Kleber has made many findings about sensory information in a musical performance. While Neuromusic research has focused primarily on the auditory modality of instrumental musicians, his work focused more on singers.

Neural Plasticity in the Corpus Callosum

Professor Kleber revisited his old dataset and performed some new analysis on the corpus callosum. As a result, the difference in the corpus callosum of the trained singers and the non-singers was proved to depend on the amount of training and the onset of training.

He made new findings as well. The ventral and dorsal regions of the sensorimotor cortex are known to activate in relation to vocal learning and motor control for speech production. What he found was that it seems to incorporate not only laryngeal functions, but also articulatory functions and exhalation.

He further wanted to see if there are any experience-dependent differences in structural connectivities between this dorsal and ventral motor cortex. The assumption being that the dorsal function is more related to vocal learning, the finding was a group difference in structural connectivity between dorsal and ventral larynx mode cortex. It means that change in structure of the laryngeal motor cortex is driven by years of singing practice.

Auditory vs Sensory Feedback in Musicians

It has also been shown that singers and musicians with experience make more use of both sensory and auditory feedback for fine motor control.

They did a pitch shifting experiment where the participants hear the pitch shift but they're supposed to ignore and continue without compensating. In terms of singing condition, non-singers responded predominantly with the auditory cortex, but the more experience they have as a singer, the more imperial and prefrontal areas were activated. It indicates some conscious control mechanism being involved.

In another experiment they looked into noise masking feedback performance. They put a lot of noise so that they have no auditory feedback and rely on something else. The result was that there is no effect on pitch accuracy in the trained singers, while untrained singers had worse pitch accuracy. The whole sensorimotor region for controlling the voice is more or less deactivated when they had no auditory feedback in the non-singers, whereas the trained singers focused more on the body and activated a network involved in sensorimotor processing and cognitive control as well as premotor and the insular cortices. It implies that singers integrated somatosensory and auditory feedback from the body when performing the task.

Innate Perceptual Differences?

At the same time, experimental results have shown that some individuals even without experience on vocal training can also make use of such information, perhaps innately. There were a couple of studies that showed that performing artists including singers, violinists, dancers and orchestra musicians appear to have better interoceptive accuracy, meaning better internal body perception. Although there is a correlation with training quantity, it is rather small.

What he finds interesting is that there seems to be a relationship with interceptive accuracy in the non-singers, and that a person who never had singing training but has better body perception can have better pitch. He thinks some of them are perhaps born with different perceptual preferences and abilities, and being able to perceive your body in a better way might help you develop your musical skills.

【Session 3 : Understanding and Managing Chronic Pain in Musician】

Guest : Professor Anna Zamorano (Aalborg University) Denmark

Professor Zamorano studies how musicians perceive and cope with chronic pain based on physiological and neurophysiological findings.

She talked about chronic pain as the persistence of or recurrence of pain that lasts more than three months. There are environmental, biological and psychological factors. We normally see that 20% of ordinary individuals suffer from chronic pain, while in professional musicians it goes up to 60-80%.

Acute Pain Perception in Healthy Musicians

They explored with heat and pain pressure the sensitivity of musicians with and without chronic pain. What they saw was that healthy or pain-free musicians had increased sensitivity just like when they get chronic pain, compared to healthy non-musicians.

According to another research they published just this year, they also found that when they provided an electrical non-susceptible stimulation, musicians showed increased response at the cortical level. The responses originated from the sensory motor areas. Perhaps it's telling us that musicians can discriminate pain better.

Acute pain perception during prolonged experimental muscle pain

They performed another experiment to learn how musicians process persistent pain. They explored the cortical responses before injecting NGF (Nerve Growth Factor), which is said to be involved in pain, in the right hand of professional musicians and explored how they were perceiving pain three days later and eight days later. They saw that only the musicians group developed higher response and sensitivity to electrical stimulation at the cortical level. It implies that in some way musical training is facilitating and changing the sensory motor areas, pain system, and the pain pathways.

Top-down control of pain in musicians

When we perceive pain we get a withdrawal response called corticomotor depression, which prevents us from having a severe injury. Normally the stronger the corticomotor suppression, the less pain is experienced. However, when pain continues for several hours, the response is reversed. The stronger the corticomotor suppression, the higher the pain gets.

They explored the corticomotor responses of the hand muscles, and saw preserved corticomotor excitability and lower pain in musicians. They saw lower functional connectivity between the insular cortex and the pain-related areas in chronic pain musicians compared to non-musicians. Musicians in general seem to modulate bodily signals that are not relevant for the performance in some way.

Experiment results have shown that musicians have not only increased or enhanced bottom-up perception, but also better top-down control mechanisms for controlling pain. Still, so many musicians still suffer pain syndromes. Her personal opinion as a researcher and clinician is that they should improve epidemiological studies. She also mentioned that training habits and strategies, beliefs and self-care routines, sleep quality, respecting pain signals, and early detection matters.

【Session 4 : Network flexibility of the brain of brass instrumentalists】

Guest : Professor Kazumasa Uehara (Toyohashi Institute of Technology), Japan

Professor Uehara has done collaboration work regarding TMS (Transcranial Magnetic Stimulation) with Sony CSL researcher Dr. Furuya for years. He presented about network flexibility of the brain of brass instrumentalists.

Exploring the neural correlates of symptoms in musician's dystonia

In the TMS experiment with pianists, he stimulated the primary motor cortex and quantified the midi data to assess the behavior feature. They found reduced SICI and the elevated ICF in dystonia pianists. They saw apparent brain activity in musicians with dystonia. Additionally, they found that reduction of SICI was associated with temporal imprecision of finger movements, while elevated ICF was associated with smooth and quick transitions of finger movement from flexion to extension. It can be said that disruption of E/I balance in M1 is associated with the loss of motor dexterity during playing in pianists with FTSD (focal task-specific dystonia).

In fMRI research, they tried to look at relationships between aberrant brain activity/somatotopy and symptoms in musician’s dystonia. They performed experiments with patients with embouchure dystonia (ED) and healthy players, and found huge instability of f0 signals in ED patients. Furthermore, ED patients showed more activity in the brain regions responsible for sensorimotor control than healthy performers. Controlling excessive brain activity may be the key to relieving dystonia symptoms.

Brain Network flexibility and skilled musical performance

Brain activity changes from moment to moment. They wanted to quantify those neurodynamics called network flexibility using a mathematical technique. They used a network matrix to assess changes of the brain network flexibility, captured network changes over time, and calculated the brain network. As a result, theta frequency was more flexible as compared to the other frequency. They also saw a relationship between the feedback control and brain network flexibility. They also found that the frontal, temporal and occipital area contributed to feedback control of tone production.

The results suggest that EEG-based brain network flexibility is capable of predicting skilled musical performance in expert brass instrumentalists. These findings propose significant potential for designing a new approach to predict an individual's skill based on neural dynamics and offer a new intervention tool, such as a neurofeedback system, to enhance effects of musical training.

Deep learning approach to detecting temporal and special features of human-human cooperative skill

They also did research on orchestra cooperation skills focusing on neural traits using deep learning and EEG data. The participants performed cooperation tasks in pairs, and recorded EEG signals from each of them. As a result, musicians turned out to be more accurate in terms of the cooperation skill compared to non-musicians. AI analysis also showed that musician’s key brain regions are active during the preparatory phase before executing a cooperative task.

As a take-home message, Professor Uehara mentioned that accurate quantification of behavioral features could help us identify neural correlates of skilled or impaired musical performance.

【Session 5 : Toward evidence-based physical education for pianists】

Guest : Dr. Shinichi Furuya (Sony CSL), Japan

Dr. Furuya's research interest is to overcome the limit of the creativity of musicians, which can be challenging sometimes due to dystonia or chronic pain, and he hopes to break that ceiling as well. As a motor control researcher, he developed a physical education program "Physical Education for Artists Curriculum (PEAC)" for artists to provide technological support and physical education. They also have discussions between scientists, engineers and artists to achieve client-tailored musical education and research.

Physical Education Curriculum Using Science and Technology : PEAC

PEAC consists of six elements, including "Lectures", which teach the principles and knowledge on the brain, body, and music, and "Coaching", which help solve musical problems with physical education in a hands-on manner. Four of the other elements were introduced as below.

  •  Lesson Support : The program provides a system that visualizes posture and finger touch using a camera that estimates the body skeletal movement and sensors embedded in piano keys and pedals. This system enables teachers to communicate nuances of expression and body use that could not be communicated verbally alone, thereby helping students overcome problems on communication of tacit knowledge.
  •  Practice Support : The program provides a sensor system and application that can accurately record and visualize touch, pedal, and body movements simultaneously. It supports creative practice by assisting in the exploration of new expressions and ways of using the body. The mobile application allows students to review the detail data regarding arm, body postures and finger movements at home to improve performance. It also helps to prevent excessive practice by recording a history of practice.
  •  Skill Check : The program visualizes students’ performance skills and assess pros and cons of students’ expertise using machine learning techniques and our sensing system. The program can identify which finger needs more sensorimotor training. For example, the program found that timing error had nothing to do with motor function, but instead found its correlation with haptic function.
  •  Advanced Training : Based on the assessment of skills, a variety of training has been developed for enhancing somatosensory and auditory skills of pianists. For example, the training uses a hand exoskeleton, a robot that moves fingers. Evidence has shown that such passive movement enhances skills, especially in younger pianists.

Gifted Brain

Dr. Furuya offers a piano academy that incorporates the above-mentioned curriculum into music education. He has also been running a longitudinal study with the gifted pianist students. They have been recording and analyzing the EEG signal, sensory and motor functions, and so on almost every month. So far research has shown that EEG signal is related to the skills of the gifted pianist, and that 1/f slope is somewhat related to the performance of the motor precisions of students.

His students in the academy have already won a lot of prizes. This is evidence proving the effectiveness of the education and research for pianists.

【Session 6 : Seven ways to prevent Musicians Dystonia】

Guest : Professor Eckart Altenmüller (Hannover University of Music, Drama, Media) Germany / ONLINE only

Professor Altenmüller has been working for 30 years to help musicians suffering from pain or focal dystonia, and came to the conclusion that it is important to prevent it.

What is Musician's dystonia

Dystonia is characterized by deterioration of fine motor control of skilled movements. It does not usually accompany pain, although pain can also develop due to overuse. It is linked so much to anxiety.

Dystonia can begin already in childhood. Genetic factors and adverse childhood events such as abuse, loss of parents and divorce of parents result in anxieties and lack of confidence, which leads to obsessive practicing and exaggerated perfectionism that causes dystonia.

Risk factors

He introduced that there are certain types of people who are more at risk of developing dystonia. Classical musicians are more at risk than Jazz musicians. With regards to gender, males tend to develop dystonic movements more, while females tend to develop chronic pain more frequently. With regards to age, individuals over age 40 are more likely to develop dystonia, and late start of musical training (after 9 years old) has higher risks as well.

Additionally, as he listened to the parents of his patients, he realized that many had terrible life stories since childhood. A questionnaire with musicians suffering from dystonia has also shown that they had more frequent adverse childhood events, including emotional abuse, higher separation from parents, emotional neglect and sexual abuse.

Prevention

Prevention is probably possible. He introduced seven ways to prevent musicians' dystonia:

  1.  When the child has a promising love of sounds and music, start early with music lessons
  2.  Loving education, respecting the autonomy and emotional needs and regulation of the child
  3.  Good teaching, supporting enhanced expectancies, autonomy and motivation of the child
  4.  Good practice behaviors, (emphasis on audiation, self-awareness, mental practice)
  5.  Respect principles of sensorimotor learning: external focus, avoid fatigue, promote good habits
  6.  Reduce triggering factors: Pain-syndromes, Psychological stresses, Bodily damage
  7.  Joy of music, good self awareness, social skills and optimistic live attitudes reduce stress

A new review could be proof that they were perhaps already successful in prevention. Since 1994, they have been collecting data of musicians from 11 centers in Germany, and the number of musicians with dystonia has been decreasing since the peak in 2013. Perhaps societal changes and the teaching in general is much better than it was in the 70s or 80s. He hopes that their education was somehow helpful for musicians' prevention of dystonia.

After the lecture, many questions were raised from the audience and gave rise to a lively discussion. The workshop concluded successfully with closing remarks by Dr. Furuya.

2023/11/27
上へ戻る