Scientists have developed a novel smartphone-based technique to diagnose viral infections that uses a deep learning algorithm to identify viruses in metal nanoparticle-labeled samples, enabling rapid virus detection without the need for skilled laboratory workers and expensive equipment. The system correctly identified clinically relevant concentrations of Zika, hepatitis B (HBV), or hepatitis C (HCV) in 134 patient samples with 98.97% sensitivity. Mobile phone subscribers are on the rise worldwide, including in sub-Saharan African populations that are heavily burdened by infection outbreaks. Since these widely available technologies also possess powerful new computing abilities and built-in sensors, scientists have identified mobile phones as a promising tool to help manage infectious diseases worldwide. To harness smartphones' virus-detecting potential, Mohamed Draz and colleagues developed a process in which samples are loaded onto microchips and labeled with platinum nanoprobes, after which a hydrogen peroxide solution is added to the chip, reacting with the platinum-nanoprobe complex to form bubbles. This bubble signal is detected using a smartphone system with a trained deep learning algorithm, which images and analyzes the bubbles to determine the microchip's viral content. Draz et al. tested this nanoparticle-enabled smartphone system using 22 HBV-spiked serum samples, 60 Zika-spiked blood samples, 27 HCV-infected patient plasma or serum samples, and 25 Zika-infected patient serum samples. Each virus sample was tested using a microchip modified to detect one specific target virus. Follow-up analyses showed that the system accurately identified all HCV-infected samples, with one false positive and one false negative for Zika-spiked samples and two false positives for HBV-spiked samples. The authors conclude that this virus detection system can be used to detect a wide range of infections and may be adapted to many different smartphone models.