Bird singing is investigated as key to develop voice prosthesis for humans

Scientists from the University of Buenos Aires discovered that the study of the neural mechanisms used by birds during singing could help to develop phonation devices for people who lost speech.

Tags Health
An Argentine research paper was selected as one of the best research papers of 2013 by Nature

Mandarin Diamond sample [Photo: www.wikimedia.org]

Human beings learn from a very young age to communicate through language imitating other humans. Similarly, certain species of birds learn their song mastered by a "tutor". Considering this similarity, scientists from the Laboratory of Dynamic Systems (LSD) of the School of Exact and Natural Sciences Natural of the UBA, discovered that the neural mechanisms used by birds during singing could be the key to develop vocal prostheses for people who lost speech

Ana Amador, Yonatan Sanz Perl and Gabriel Mindlin from the LSD together with Daniel Margoliash from the University of Chicago, discovered that dynamics of bird singing is encoded in a group of neurons in the premotor cortex, which constitute a nucleus called Higher Vocal Center (HVC), acting predictively and in connection with different brain structures related to song production and learning.

This research debunks the old view that sustained that a group of cells would indicate how and when the various muscles and elements involved in singing should intervene. "Similarly to humans, birds need a tutor to learn how to sing, i.e. someone to copy, and the HVC structure is involved at that moment" explained Ana Amador, CONICET postdoctoral scholar in LSD.

HVC cells will configure for incorporating learning information necessary for bird singing acting simultaneously with singing muscles. According to Gabriel Mindlin, CONICET main researcher and lab director: "understanding the physical mechanisms involved in singing and translating them into simple mathematical equations allowed us to approach tiny electronic devices, able to imitate a complex behavior such as voice and open a new chapter in the vocal bioprosthetic".

Sensors were put in the vocal tract and muscles of a Diamond Mandarin sample and it was temporarily silenced to understand the neural mechanisms of birds. Muscles activated, pressure exerted in the air sacs and neuronal mechanisms were registered when the bird tried to sing. The information obtained allowed them to identify the most important factors for phonation and developed a simple mathematical model that allowed them to produce the song of a bird synthetically and in real time.

Scientists at the laboratory of the School of Exact Sciences transferred the nonlinear model to an artificial synthesizer or chip that when receiving the movement of the muscles of the bird’s vocal apparatus issued a sequence of electronic pulses, which when translated into sound, resulted in the song of a bird in real time. This work, with which Ezequiel Arneodo conducted his doctoral thesis represents a scientific milestone unique of its kind because it related to the first miniaturized device, i.e., an electronic syrinx, which synthesizes a bird sing in real time.

The resulting synthetic singing is so similar to the actual singing that birds confuse it with their own signing. Thus, Ana Amador confirmed from the United States, where she was conducting his post doctorate that the neuronal activity of birds activated with her own voice acted similarly before the synthetic sound. This breakthrough was presented at the annual conference of the U.S. Society for Neuroscience.

Everything shows that if this line of research continues it would be possible to create a device to be applied on human beings who have lost speech. "If we understand in the same way how the human the vocal apparatus, the vocal cords and the upper vocal tract work, and how the brain sends motor instructions and how they produce the diversity of elements that make up human speech, then we could apply to a portable prosthetic solution at a very low computational cost" Mindlin says.

Unlike birds, the human vocal apparatus has a sophisticated modulation of lips and tongue. The challenge in this case, is to achieve a noninvasive measurement of the minimum number of physiological, muscle and pressure-related parameters required for operating the chip and later, synthesize a voice accurately, to allow people who lost speech to communicate.

This research carried out by Ana Amador, Yonatan Sanz Perl, Gabriel Mindlin and Daniel Margoliash about bird singing was published in the Revista Nature in March 2013. Editors of the prestigious journal selected this article as one of the best articles published in the year.

The Ministry of Science, Technology and Productive Innovation granted to the Laboratory of Dynamic Systems a subsidy for the amount of $ 635,700 through the Fund for Scientific and Technological Research (FONCyT) of the National Agency for Science and Technology Promotion for the research development.

Sample Cases
  • An Argentine research paper was selected as one of the best research papers of 2013 by Nature
Image of
An Argentine research paper was selected as one of the best research papers of 2013 by Nature

Mandarin Diamond sample [Photo: www.wikimedia.org]

Video of