Breathable soft bioelectronics for enhanced automatic detection of obstructive sleep apnea.

Obstructive sleep apnea (OSA) is a common sleep disorder that affects 1-5 % of children, with a notably higher prevalence among those with cleft lip/palate and dentofacial deformities. Early diagnosis in these populations is essential; however, access to polysomnography is often limited due to its high cost, technical complexity, and associated discomfort. Here, we introduce a wireless, soft, and breathable bioelectronic system for detecting OSA. This wearable device features a perforated, deformable structure that improves skin conformity, facilitates sweat removal, and minimizes motion artifacts. By measuring electrophysiological signals on the face, the system employs a deep learning framework that integrates multi-stream convolutional neural networks with bi-directional long short-term memory models to classify sleep stages and detect apnea events automatically. This work establishes a foundation for the diagnosis of pediatric OSA and has the potential to differentiate pre- and post-operative sleep patterns in patients undergoing orthognathic surgery.