AL-powered camera system for medication error detection in real-time at the point of care

Active Grant Awarded

Award ID
R41EB039615
Recipient
Awarding agency
Obligated
$349,828

Award facts

Completion date 20% complete
Place of performance
Roanoke, TX, USA

Award details come from USAspending.gov.

About this award

The Department of Health and Human Services' National Institutes of Health, through the National Institute of Biomedical Imaging and Bioengineering, awarded Vigilant Labels Inc. $349,828 on September 9, 2026, under the Discovery and Applied Research for Technological Innovations to Improve Human Health program (CFDA 93.286). Grant R41EB039615 funds development of Verify Vision, a camera-based, edge-AI system that detects vials and syringes in a provider's hands, decodes the drug label, and cross-checks vial and syringe identity before injection, surfacing an audiovisual alert only when needed. The system would transmit recognized syringe contents to the electronic medical record, enabling patient-specific safety alerts before the syringe is pushed. Existing EMR drug-interaction and allergy modules generate warnings only when they know in advance which medication the clinician intends to give, and this capability addresses that gap. The project aims to advance the technology for real-time detection in a true clinical environment. Specific Aim 1 determines the optimal data-collection modality that ensures high accuracy while maintaining clinical staff acceptability, testing new modalities including a higher-resolution option.

The period of performance runs from September 9, 2026 to an ultimate completion date of September 9, 2027, with work performed in Roanoke, Texas. The assistance type is a grant, and the program supports hypothesis-, design-, technology-, or device-driven research on technologies for biomedical imaging and bioengineering.

Description

Abstract
Vigilant is developing an AI-powered camera system for medication error detection in real-time at the point of care. Medication-related errors are a major contributor to these adverse events, with up to 12% of errors leading to serious injury or death. Within anesthesia, drug administration errors are the most commonly reported critical incidents, while in intensive care they are a leading cause of serious medical errors. Adverse events from injectable medications cause 1.2 million hospitalizations annually, amounting to $5.1B in healthcare costs. Vigilant is developing verify vision, a camera-based, edge-AI system that detects vials and syringes in the provider's hands, decodes the drug label, and cross-checks vial and syringe identity before injection-surfacing an audiovisual alert only when needed. While enterprise electronic medical records (EMRS) already contain drug-interaction checking and allergy modules, they can only generate warnings if they know in advance which medication the clinician intends to administer. Verify vision fills this GAP by automatically recognizing the syringe contents in real time and transmitting that information to the EMR. This feature would enable immediate, patient- specific safety alerts before the syringe is pushed. The approach uses deep learning to automatically detect syringes and vials in a provider's hand, classify the drug type on the label, and automatically check if they match in order to detect vial swap errors. In this project, we seek to advance our technology to be suitable for real-time detection in a true clinical environment. To accomplish this we propose the following specific aims: aim 1: determine the optimal modality for data collection that will ensure high levels of accuracy while maintaining high levels of acceptability among clinical staff. This project will test new modalities including (1) a higher resolution camera, (2) a fixed position camera , (2) a hybrid approach of both modalities at once. Aim 2: implement software improvements including improved detection accuracy, object tracking, and processing efficiency to support real- time detection. IT will support real-time implementation of our system including training with diverse and synthetic data, implementing multi-object tracking with occlusion handling, utilizing newer versions of yolo and advanced detectors, optimizing field of view and cropping, and model pruning and edge AI optimization. With these improvements we aim to increase detection accuracy and reduce system latency. Aim 3: demonstrate accuracy of system to discern error and non-errors in a mock clinical environment. The system's ability to detect these errors in real time will be tested, as well as provide generate explanatory reasoning for each flagged error. Phase I will produce the evidence and technical artifacts to justify phase II scale-up across or/ICU suites and to begin regulatory planning for software as a medical device.

Quick answers

Who received award R41EB039615?

R41EB039615 was awarded to Vigilant Labels Inc. by National Institute of Biomedical Imaging and Bioenginering (National Institutes of Health) on September 9, 2026.

How much is award R41EB039615 worth?

$349,828 has been obligated on R41EB039615.

On GovTribe

See this award's full record

  • Every transaction and subaward
  • The recipient's other awards and teaming partners
  • Similar grant awards
  • Ask GovTribe AI about this award