Alzheimer’s early detection: Language test to detect dementia in 60 seconds

Dementia screening in just 60 seconds: AI-based rapid test could soon be available in medical practices

Alzheimer’s and other dementia disorders are on the rise worldwide. There is no known cure, and the search for the causes is still the subject of research. The amyloid hypothesis is still under scrutiny, especially since the once groundbreaking study from 2006 on this topic has just been withdrawn1, new drugs such as Lecanemab are criticized for their high side effects and the enormous organizational effort (a marketing authorization decision by the EMA for Europe is expected in 2025) and the physical therapy approaches that have been shown to delay these diseases and significantly improve the quality of life of those affected, such as Transcranial Pulse Stimulation (TPS) from the field of non-invasive brain stimulation (NIBS) are still not widespread enough to have a widespread impact on this neurodegenerative pandemic of our time.

All this is also reflected in the dramatically increasing number of people in need of care, a considerable proportion of whom are Alzheimer’s and other dementia patients. In Germany, for example, people were shocked at the end of May 2024 when Federal Health Minister Karl Lauterbach presented new figures: 360,000 instead of only 50,000 new cases of care were added in the Federal Republic of Germany in 2023 alone – this figure was “almost explosive”, according to Lauterbach, and there was now an “acute problem”.2

More early detection and prevention: ways to limit dementia at an early stage

Early diagnosis and preventive measures to avoid Alzheimer’s or other forms of dementia are therefore urgently needed. In addition to the development of further blood tests (see also: ), Hungarian researchers have now presented a simple language test that is said to provide clear indications within just 60 seconds as to whether a patient is suspected of already suffering from dementia or of developing such a condition in the near future.

The focus of this test is not on an exact diagnosis, but rather on enabling general practitioners, who only have a few minutes per patient, to decide whether the person being tested should be referred directly to a specialist for more extensive dementia tests. This could save valuable time. In addition, the test should encourage patients who respond positively to the test to address the issue of prevention at an early stage (see also ).

Novel speech analysis for the early detection of cognitive impairment

The analysis of temporal anomalies in speech production could therefore prove to be particularly suitable for quickly identifying mild cognitive impairment and determining the risk of future dementia. These findings were recently presented at the 32nd European Congress of Psychiatry.

Psychiatrist Dr. János Kálmán and his research team at the University of Szeged in Hungary have developed the automated speech analysis procedure, the so-called Speech-Gap-Test (S-GAP-Test), over the past five years3,4. This test analyzes temporal peculiarities in speech production. According to Kálmán, the test is characterized by the fact that it is not unnecessarily complicated by phonological and semantic aspects of language.

Speech-gap test: analysis of temporal effects of speech

The speech gap test analyzes various parameters of speech production, including speaking rate, frequency and duration of pauses, and hesitation. The research has shown that a 60-second speech sample is sufficient to obtain meaningful results. In contrast, other automatic speech analysis programs under development require significantly longer speech samples and thus take more time.

After testing various approaches, the research team decided to analyze temporal speech parameters: “These parameters are less dependent on cultural and educational influences and could therefore provide more reliable results than semantic analyses,” says Kálmán.

Promising results across language barriers

Although the S-GAP test was originally developed with Hungarian native speakers, Dr. Kálmán’s research team has since shown that the test is equally effective with English and German native speakers. The next step is to validate the test with Spanish native speakers.

An efficient instrument for the early screening of dementia, but not for diagnosis

The S-GAP test is currently designed exclusively for early screening and is not intended for the diagnosis of dementia. Dr. Kálmán emphasizes that his team therefore does not intend to register the test as a medical device.

The research team’s objective has now shifted from a purely telemedicine solution to a broader application that could soon be available to everyone. The long-term goal is to make the S-GAP test available as an app that can be downloaded and used on a smartphone. Dr. Kálmán also does not rule out the development of a more sophisticated version of the S-GAP test that could combine temporal speech parameters with biomarkers or other clinical parameters. Such a test could then also help neurologists in the diagnosis of Alzheimer’s disease and other dementias.