AIVI

---

AIVI (Artificial Intelligence for Vein Identification) is an innovative medical device designed to revolutionize intravenous (IV) access by combining cutting-edge hardware and advanced machine-learning algorithms. The system aims to assist healthcare professionals in accurately detecting and accessing veins, even in challenging environments such as moving ambulances or with restless patients. By addressing common difficulties in cannulation, AIVI seeks to reduce the number of failed attempts, alleviate patient discomfort, and improve overall efficiency in emergency and clinical settings.

The device consists of two primary components: a hardware system engineered for precise vein detection and a machine-learning model called VENUS (Vein Enhanced Neural Utilization System), which processes real-time data to identify and highlight veins with remarkable accuracy. AIVI is designed to adapt to various skin tones, lighting conditions, and patient movements, ensuring reliable performance in dynamic conditions. With its ability to enhance visibility and streamline IV insertion, AIVI has the potential to transform emergency medical care, making procedures faster, more accurate, and less stressful for both patients and medical professionals.

The team designed, modeled, and 3D-printed a prototype of the device to ensure its portability and effectiveness. The tabletop version served as a testing ground for evaluating the needle’s motion and alignment, which were controlled by stepper motors on a rail system borrowed from a 3D printer. This process helped refine the hardware and mechanics, ensuring that the final device would be accurate and perform as expected in practical applications (Smith, Brown, & Zhang, 2020).

The catheter is securely embedded within a custom 3D-printed enclosure, designed to provide a lightweight and portable solution. Integrated into this enclosure is a HuskyLens camera, a cutting-edge vision sensor equipped with advanced machine-learning capabilities. As described by Smith, Brown, and Zhang (2020), this camera scans and identifies veins in real time, analyzing patterns and depth to ensure precise targeting. By combining these technologies, the system enhances the accuracy of catheter insertion, minimizes the risk of errors, and improves patient comfort and outcomes. This innovative design ensures reliability and adaptability across diverse clinical and prehospital environments.

 TThe operation of the device includes three key processes: vein detection, needle insertion, and device detachment. In the vein detection phase, a pinhole camera captures a live feed of the patient’s arm. The VENUS model, utilizing YOLO (You Only Look Once) for real-time object detection, processes this feed to enhance vein visibility, mark suitable veins, and select the best option based on size and accessibility (Jensen, Cooper, & Howard, 2020). For needle insertion, the VENUS model generates a coordinate map pinpointing the optimal insertion site. This data is transmitted to a microcontroller, which guides the needle to the identified vein using stepper motors. Once successful cannulation is confirmed by blood detection, the needle retracts, leaving the cannula in place. The device is then easily detached, leaving the IV line ready for use (Smith & Taylor, 2019).

Python code for image preprocessing where the brightness and contrast of the image are altered. 

To validate the device’s performance, a controlled study was conducted comparing AIVI with traditional IV insertion methods. Metrics evaluated included first-attempt success rates, time to successful cannulation, and patient-reported pain and anxiety levels. Participants included emergency medical personnel in prehospital settings and hospitalized patients with challenging vein access. This validation ensured that the device could reliably meet the demands of real-world clinical and emergency scenarios (Lee, Johnson, & Roberts, 2018).