Dysphagia, prevalent among Parkinson's and stroke patients, hinders proper eating, impacting their quality of life and potentially leading to fatal outcomes if untreated. Currently, Videofluoroscopic Swallowing Study (VFSS) is the gold standard for diagnosis but requires specialized facilities and trained staff. While many wearable devices have been developed to ease these burdens, none could reliably detect specific dysfunctions like silent aspiration without VFSS. We present a multimodal wearable swallowing monitor incorporating machine learning for automatic dysfunction assessment and silent aspiration diagnosis. The device, featuring a kirigami pattern, is directly mounted on the neck for continuous, high-fidelity monitoring of electromyograms and swallowing sounds. The built-in machine learning algorithm classifies various swallowing patterns, including silent aspiration. Clinical trials with stroke patients underscored the device's significance, matching the VFSS in detecting swallowing disorders. This wearable technology holds promise for advancing dysphagia healthcare and post-stroke rehabilitation therapy.
|