AI-Powered Specimen Tube Verification Prototype

In busy outpatient clinics, specimen collection involves high volumes of manual tube selection and label checks, placing significant demands on staff knowledge and attention. To better support our staff and enhance patient safety, we developed an AI-powered verification prototype that serves as a reliable second check at the point of specimen collection. This first-of-kind standalone prototype integrates Optical Character Recognition (OCR) for label reading, Computer Vision for tube recognition, and a label-tube matching algorithm that generates an LED match/mismatch indication — empowering staff with real-time verification support within their existing workflow. Its standalone plug-and-use design is compact and deployable without IT integration, and users simply place the label and tube to receive instant feedback. Validated at 97.4% accuracy, with zero specimen errors recorded during clinical deployment.