Start-up Accexible is using artificial intelligence (AI) to assess the linguistic content of a person’s speech as well as how they are speaking to identify potential underlying health problems, such as Alzheimer’s, depression and anxiety. The technology, based on a combination of neuroscience, linguistics and mathematics, is accessible via phone call, computer or app and is designed to be used by general practitioners as a screening test and neurologists as a monitoring tool. The company claims 90% accuracy and results in just a few minutes. The service is already being used by 10,000 patients and 40 practitioners across five markets and trials are being run in Spain, Colombia and the UK.
A preclinical stage of Alzheimer’s can appear without tangible symptoms, but the appearance of certain biological biomarkers may indicate the disease. Accexible is exploring whether the model can predict levels of beta amyloid, the protein that builds up inside the brain in Alzheimer’s patients. The company’s CEO, Carla Zaldua Aguirre says she hopes that the system will ultimately detect a range of health conditions and expedite access to treatment.
These types of healthcare solutions are hitting the market in increasing numbers as the potential benefits of using AI become more widely known. In January, researchers from the Regenstrief Institute and Indiana University School of Informatics in Indianapolis argued that the analysis of speech coordinations and movements could lead to earlier diagnoses of Parkinson’s. Researchers from Mayo Clinic in Arizona have successfully used AI in analysing records from 600 stomach cancer patients to identify the lead factors in the onset of the disease. The technology was better at identifying high-risk patients than existing models, according to the study published last February. There is also a growing market for voice-activated technology in the field of health, with Amazon’s Alexa in trials in several parts of the US. Even smaller startups are contributing to the growing buzz, such as Babylon Health which raised $60m last year to develop its health app which features an AI doctor. Competition in this field is expected to increase over the coming years as more firms develop healthcare solutions that rely on AI to reduce error rates and workloads, as well as facilitating earlier diagnoses of diseases.