An interesting proof-of-concept study by researchers at the University of California San Diego School of Medicine reveals that AI tools for analyzing speech can effectively predict the level of loneliness in older adults. Loneliness and its consequences constitute a relevant public health problem: the WHO warns that cardiovascular diseases are the main cause of mortality, and in this sense there are studies that associate loneliness is associated with higher blood pressure figures, alterations of the system immune system and increased risk of premature death, in addition to depression, anxiety and sedentary lifestyle. So much so that unwanted loneliness is one of the most prevalent fears, especially among the generation of babyboomers over 65, who are socially, ethically and aesthetically invisible. 9 out of 10 want to grow old at home and with company, but few succeed: in the contemporary world, loneliness is another of the pandemics. Without going any further, in 2018 a Ministry for Solitude was created in the United Kingdom with Minister Tracey Crouch. How can we detect loneliness more effectively? Using Artificial Intelligence. HBO Spain is giving away 14 days of free subscription with no commitment of permanence to all its new users. You can cancel whenever you want or keep the subscription for only € 8.99 per month. NLP tools to detect loneliness Natural language processing (NLP) consists of a series of techniques that process or analyze large volumes of speech and unstructured natural text, with the help of AI and machine learning systems. Fascinating early studies suggest that conditions such as psychosis, post-traumatic stress disorder, bipolar disorder, and depression can be detected by analyzing a person’s natural speech. NLP tools could be applied to the detection of loneliness, a growing health problem described as a more important factor of premature mortality than obesity. Ellen Lee, lead author of the new research, suggests that loneliness is a particularly difficult psychiatric condition to measure, and because clinicians often struggle to quantify loneliness in patients, objective measures are urgently sought. Since many studies ask direct questions about the frequency of loneliness that can lead to biased responses, this project decided to use natural language processing, or NLP, an unbiased quantitative assessment of expressed emotion and feeling, along with the common tools for measuring loneliness. For this, they recruited 80 older adults, who in addition to being evaluated with conventional instruments, completed a conversational and semi-structured interview lasting an hour and a half. The interviews were transcribed and later analyzed using a natural language system developed by IBM. In addition to detecting loneliness in subjects not captured by conventional assessments, the system discovered differences in the way men and women talk about loneliness. “NLP and machine learning allow us to systematically examine long interviews of many people and explore how subtle features of speech, such as emotions, can indicate loneliness,” says Varsha Badal, first author of the study. The system was able to qualitatively predict a subject’s loneliness with 94% accuracy. Notably, the more lonely a person felt, the longer were their responses to direct questions about loneliness. The researchers even suggest that the presence of a kind of “lonely speech” pattern could be used in the future to monitor the well-being of older subjects. Men used more words related to fear or joy, while women were more likely to explicitly express feelings of loneliness. The next stage of the investigation will be to combine other sensor data in the assessments (such as GPS tracking and sleep data) to personalize each individual finding. In addition, the system will need to be tested on larger and more diverse populations to fine-tune and optimize its accuracy. “Over time, complex AI systems could intervene in real time to help people reduce their loneliness by embracing positive cognitions, controlling social anxiety, and participating in meaningful social activities,” the research concludes. The study was published in The American Journal of Geriatric Psychiatry. This article was published on TICbeat by Andrea Núñez-Torrón Stock.