+1 Recommend
1 collections
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Further developing voice analysis technology capable of detecting changes in vocal patterns which signal changes in stress levels and depressive symptoms


      Science Impact, Ltd.

      Read this article at

          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.


          In situations of catastrophe it is impossible for specialists to conduct interviews, diagnose and treat all the affected individuals. However, similar conditions exist in society as well. In large corporations for example, it is extremely time consuming to identify and treat those who are feeling the effects of stress, anxiety or depression. The logistics are staggering and compounded by the aforementioned issues of reporting biases. 'Therefore, a screening technique which is inexpensive, simple and highly accurate is necessary,' says Tokuno. His solution relies on the fact that stress exerts an impact on human emotions. According to Tokuno: 'We developed a method to measure the mental health of a speaker based on the variations in emotional components extracted from the voice, rather than directly analysing stress conditions based on vocal data.' Tokuno and his colleagues have shown that the technology works. Several studies have been conducted showing the accuracy and reliability of the method. 'The main advantages of the present technology are its ease of implementation and low cost,' says Tokuno. 'Its application in preventative medicine is therefore promising, and, if used in combination with the various biomarker-based diagnostic techniques currently in development, better induction to healthcare specialists will be possible.' The next steps will be fine-tuning the software and implementing it into a marketable package, such as a smartphone app. The development of high-accuracy parameters for the software requires a great deal of speech data that is further linked to medical data, something no one institution could do on their own. Then there is the issue of language. 'For instance, it would be theoretically possible to use the present technology for languages other than Japanese, but no evidence currently exists, and verification for other languages cannot reasonably be done in Japan,' explains Tokuno. To address this, a joint research infrastructure using cloud technology to collect a variety of speech data is being built. Condensing the technology onto a smartphone is also in future plans. Tokuno and his colleagues envision an app called MIMOSYS: Mind Monitoring System, in which the users' vitality and mental activity are converted into a numerical value, which is displayed in a number of easy to read displays or charts. According to Tokuno, a large-scale demonstration experiment is underway using a prototype of this app. In the future, they believe MIMOSYS could be commonplace and used to diagnose many conditions, including Parkinson's disease and dementia. Tokuna says that results for depression, dementia and Parkinson's have all been promising and a plan to expand the technology for schizophrenia, autism and suicide prevention is underway.

          Related collections

          Author and article information

          Science Impact, Ltd.
          June 05 2019
          June 05 2019
          : 2019
          : 6
          : 43-45
          © 2019

          This work is licensed under a Creative Commons Attribution 4.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

          Earth & Environmental sciences, Medicine, Computer science, Agriculture, Engineering


          Comment on this article