Lithuanian scientists are developing technology for more efficient screening of voice pathologies

Lithuanian scientists are developing technology for more efficient screening of voice pathologies

For advanced throat cancer, a complex surgery called a laryngectomy is performed, which involves removing the organ vital to a person's ability to speak. In order to improve the patient's quality of life after laryngectomy, a team of software engineers from KTU, led by researcher Rytis Maskeliūnas, and a team of medical professionals from LSMU, led by Prof. Virgilijus Ulozas, conducted a study in which they looked for pathologies in patients' voices using artificial intelligence (AI).

After surgery, the patient breathes through an opening in the windpipe in the neck called a tracheostomy. Although the tracheal hole in the neck allows the patient to breathe after a laryngectomy, it irrevocably changes the patient's everyday life.

Rytis Maskeliūnas shares his experiences with voice changes after laryngectomy, which often depend on the severity of the situation. “Some people’s voices change slightly, others speak like robots, others wheeze,” says the scientist.

Helping people after surgery

According to doctors at LSMU, a very common reason for surgery is cancer, which is often caused by smoking and alcohol consumption.

R. Maskeliūnas shares that the idea of ​​studying the voice quality of patients after laryngectomy was developed by a team of doctors led by an expert in the field, Prof. Dr. Virgilius Ulozas.

To optimally facilitate the postoperative process, modern AI technologies have been used, which make it possible to reduce the workload of a specialist and eliminate the need for self-diagnosis or the patient filling out questionnaires. According to scientists, due to this research, the screening process after surgery will be more automated and therefore it will be easier to track the progression of the disease and diagnose patients.

It's like a kind of screening tool to possibly suspect that the patient has certain vocal pathologies. As a rule, the doctor analyzes the voice and creates an index of impairment, and the patients also fill out a questionnaire themselves about how they perceive the quality of the voice. Digital signal analysis additionally uses speech signal energy, formants and other parameters.”

Rytis Maskeliūnas, researcher, LSMU

Artificial intelligence algorithm

"We assume that the AI ​​algorithm used in this research may allow us to check and evaluate various voice parameters and then assign them to a specific class. It can also become a tool for tracking how a person heals, how his voice changes, in a good or bad sense," emphasizes R. Maskeliūnas.

This study is special, according to the researcher: “Previous studies have never fully utilized artificial intelligence methods as an “expert assistant” for voice analysis, and only a low level of automation has been applied in regional medical IT practice for this area.” It is important to mention that similar technology has already been used by researchers to study the voice of Parkinson's and other diseases in practice.

To determine the voice index, a smartphone or other device with an internet connection is sufficient. After uploading an audio recording or the live conversation, the system displays an estimate, on the basis of which the further course of treatment is decided and changes in the voice are analyzed.

“I am happy to be part of this multidisciplinary team led by Prof. Virgilijus Ulozas, and I really enjoy working with doctors Kipras Pribuišis, Evaldas Padervinskis and the others,” says R. Maskeliūnas.

The research solutions are clinically tested in the clinics of the Lithuanian University of Health Sciences Kaunas Hospital.

Source:

Kaunas University of Technology (KTU)

.