Artificial Intelligence-Supported Colorimetric Multibiomarker Sensor to Enable Critical Neonatal Monitoring

Clinical monitoring in the most vulnerable patients, such as newborns, relies on invasive and costly procedures and/or wired sensor surveillance, increasing discomfort and risk for undetected events. Addressing this critical need, we present a noninvasive, AI-supported multibiomarker monitoring sensor capturing multiple critical body functions via colorimetric analysis of body fluids. The sensor is optimized for its use in critically ill preterm neonates, where transepidermal fluid mirrors the interstitial compartment, thereby opening new avenues for future clinical monitoring. The silk-based sensor introduces twelve different colorimetric inks that stabilize labile bioresponsive molecules to a simple, wax-printed paper microfluidic wearable patch to facilitate real-time tracking of biomarkers (temperature (T), pH, sodium (S), and glucose (G)) on a miniaturized surface. Colorimetric responses showed high precision and reproducibility in sensing ranges relevant for neonatal care (T: 32-41 [°C]; pH: 3-9; S: 2.92-29.20 [mg/mL], G: 0.039-0.625 [mg/mL]). Deep learning for color response quantification achieved high-precision estimates (T: 0.455 [°C]; pH: 0.416; S: 0.857 [mg/mL], G:0.019 [mg/mL]; mean absolute error). The sensor's optimized interface for sample collection on skin and its performance under clinically relevant conditions, e.g., increased humidity (80%), tracking on moving object (0.986 AP@IoU = 0.5), sensor shear/rotation, uncontrolled light, as well as successful capture of physiological effects in biological fluids together with its miniaturized design caters to the critical needs of intensified monitoring in high-end patient populations such as neonates.