A team at the Icahn School of Medicine at Mount Sinai has trained an AI model on wearable signals. Its target: the long sitting spells that women with chronic pelvic pain often slide into.
The results appeared September 30 in npj Women’s Health. They sketch a future in which a digital tool nudges someone to move at the right moment instead of scolding them afterward.
The condition touches about one in seven women. It frequently travels with endometriosis, adenomyosis or fibroids. When pain and fatigue take hold, sitting for hours becomes the norm. Those same stretches then worsen the symptoms in return.
Generic coaching to sit less rarely fits that reality. So the researchers trained a forecasting model on data logged over time by the wearables these women already used. It predicts when a sedentary stretch is likely during waking hours, leaving time for a brief prompt to stand or take a short walk.
Senior author Ipek Ensari, an assistant professor of AI and human health at the Icahn School of Medicine and a member of the Hasso Plattner Institute of Digital Health at Mount Sinai, said the aim was to move past after-the-fact coaching. She wants to reach people before inactivity begins, with a nudge short enough to fit an ordinary day.
