BUDAPEST — Analyzing temporal changes in people's speech could be a simple way of detecting mild cognitive impairment to see whether there is a risk of developing dementia in the future, suggests research.
János Kálmán, MD, PhD, and colleagues at the University of Szeged in Hungary have developed an automated speech analysis approach called the Speech-Gap Test (S-GAP Test) that is unique because it focuses on the temporal changes made when someone talks. This means it does not overcomplicate matters by also assessing the phonetics and semantics of speech, Kálmán told Medscape Medical News.
Kálmán presented his findings at the 32nd European Congress of Psychiatry.
Temporal Speech Parameters
The test analyzes parameters such as how quickly someone speaks, whether they hesitate when they talk, how long the hesitation lasts, and how many silent pauses they make. This can be done with a mere 60-second sample of speech, Kálmán said, noting that other automated speech and language tools currently in development need much longer audio samples.
"We tried different approaches and we finally ended up with the temporal speech parameters because these are not culture-dependent, not education-dependent, and could be more reliable than the semantic parts of [speech] analysis," he explained.
The analysis of temporal speech parameters is also not language-dependent. Although the S-GAP Test was developed using audio samples from native Hungarian speakers, Kálmán and his collaborators have shown that it works just as well with samples from native English and German speakers. They now plan to validate the test further using samples from native Spanish speakers.
For Screening, Not Diagnosis
Currently, "the only purpose of this tool would be initial screening," Kálmán said at the congress. It is not for diagnosis, and there is no intention to get it registered as a medical device.
A national survey of primary care physicians conducted by Kálmán and collaborators showed that there was little time for performing standard cognitive tests during the average consultation. Thus, the original idea was that the S-GAP Test would be an aid to help primary care physicians quickly flag whether a patient might have cognitive problems that needed further assessment at a memory clinic or by more specialist neurology services.
The goalposts have since been moved, from developing a pure telemedicine solution to a more widespread application that perhaps anyone could buy and download from the internet or using a smartphone.
Kálmán doesn't discount developing a more sophisticated version of the S-GAP Test in the future that combines temporal speech parameters with biomarkers for mild cognitive impairment or Alzheimer's disease and could be used in memory clinics and by neurologists for the purpose of diagnosis.
Detect to Prevent?
The big question is: What happens to all the people that could be flagged as needing further assessment using tools such as the S-GAP Test?
Tackling risk factors for dementia will probably be key, said Robert Perneckzy, MD, MBA, professor of translational dementia research at Ludwig Maximilian University of Munich, Imperial College London, and the University of Sheffield.
According to the Lancet Commission on dementia, there are 12 potentially modifiable risk factors for dementia. Their influence varies throughout the life course, but certain early events, such as the level of education attained, can't be modified in an older patient.
That said, there are many risk factors that might still be influenced later in life, such as adequately treating comorbid conditions such as diabetes, and addressing alcohol consumption and smoking practices.
"We can do things in terms of personal risk, risk mitigation, which have a huge effect on dementia risk much later in life," said Perneckzy.
"The speech-based assessments are another opportunity to save our time as doctors to do assessments before someone comes to the memory clinic," he said.
The S-GAP Test is under development by the University of Szeged. Kálmán is a co-inventor. Perneckzy had no relevant conflicts of interest but has helped validate the S-GAP Test in the German language.
Sara Freeman is a freelance medical journalist based in London, England.