Newly, the doctor goes to the patient's bedside with a microphone connected to a smartphone or tablet with a special application and records the conversation. It will then be evaluated by artificial intelligence - it will help determine the extent of speech damage and suggest further treatment. "The investigation will be shortened three times. Artificial intelligence will not do the work of doctors, but will help them. The speech-language pathologists will, of course, choose the next procedure. Jiří Mekyska is developing the device with colleagues from the Institute of Telecommunications, within the university's startup company Scicake.
The team's goal is to collect enough recordings of speech, from normal to very bad speech, that the artificial intelligence can learn from. In time, it will be able to determine by itself how extensive the disability is when recording the patient. The speech therapist thus obtains independent supporting data for treatment.
The result should not only be a faster and better diagnosis, but also a faster way to correct the patient's speech so that he can return to life as soon as possible. Our device will be a valuable helper, choosing the right treatment and subsequent individual therapy will always be up to the speech therapist. So: we are not trying to replace him, but to help him.
Created by the team of Czech Trade Scandinavia
Source: https://www.novinky.cz/clanek/veda-skoly-umela-inteligence-z-vut-pomuze-logopedum-s-lecbou-poruchy-reci-40451778