An acute hearing at any age: artificial intelligence will help doctors make a diagnosis

An acute hearing at any age: artificial intelligence will help doctors make a diagnosis

Researchers at ETU “LETI” presented the results of research aimed at solving the problem of hearing impairment in the elderly to the European scientific community.

05.06.2019 146

The 53rd Annual Meeting of the European Society for Clinical Investigation 2019 took place in Coimbra, Portugal, from 22 to 24 May. Oleg Markelov, Associate Professor of the Department of Radio Engineering Systems, presented the results of research carried out at the Department of Automation and Control Processes and the Department of Radio Engineering Systems of ETU "LETI" together with the Pavlov First Saint Petersburg State Medical University and the University of Oldenburg (Germany).

Birger Kollmeier, President of the European Federation of Audiology Societies, Professor at the University of Oldenburg, headed the research group from the German side. The medical part of the study from the Russian side was carried out by a research group under the direction of Maria Boboshko, Professor, Head of the Laboratory of Hearing and Speech at Pavlov First Saint Petersburg State Medical University.

The purpose of the studies is to solve one of the most common age-related problems - hearing impairment in the elderly. According to WHO, 30% of people suffer from hearing loss after 60 years and 2/3 of people after 70 years. Researchers conducted an intellectual analysis of the results of audiological tests in people of different ages, provided by the German and Russian sides.

“Now, we continue the research. We have applied for a joint grant of the Russian Foundation for Basic Research and the German Research Foundation. We plan to create an expert diagnostic system that recommends treatment and a rehabilitation program based on the results of audiological and speech tests will diagnose regardless of the language in which the study was conducted.”

Oleg Markelov, Project Manager from the Russian side, Associate Professor of the Department of Radio Engineering Systems of the ETU “LETI”

The research methodology is based on data mining, machine learning, and statistical analysis devices. The system will be trained on the pre-processed data of the results of audiological tests, both from the Russian and from the German side. The test suite is constantly expanding.

According to leading Russian audiologists, there are no such systems on the domestic market today. In Germany, under the leadership of Birger Kollmeier, a prototype of the system was developed under the name CAFPA but, it is less functional since the test suite and languages are very limited.

“Doctors highly appreciate the results of the research, because at the moment there are no systems that perform automatic analysis of the results of audiological tests in Russian clinical practice. In addition, the system will automatically make a diagnosis based on a test suite, which will seriously reduce the likelihood of an error.”

Maria Boboshko, Head of the Laboratory of Hearing and Speech at Pavlov First Saint Petersburg State Medical University

A number of researchers and practicing audiologists from Europe, who highly appreciated the scientific potential of the study and offered to expand the language coverage of tests, became interested in the results of the study. In particular, researchers discuss the possibility of including the analysis of test results in English, French, Spanish, and Portuguese.

The research results were published in the European Journal of Clinical Medical Research.