Vein Identification and Localization for Automated Intravenous Drug Delivery System

Intravenous (IV) drug administration of fluids or medication and blood drawing procedures require the medical practitioner to access the veins in patients. This procedure is of prime importance in saving an injured soldiers life on the battlefield by intravenously administering haemostatic agents to prevent blood loss due to haemorrhage. This paper presents an image acquisition system that captures images of subcutaneous veinpatterns and processes these to conclude the location for intravenous needle-insertion. The system consists of a camera, receptive to near-infrared light, which captures an image of the subcutaneous veins illuminated by a suitable source of light. The contrast-enhanced image of the veins thus obtained is processed using an algorithm to conclude an ideal site for needle-insertion. Subcutaneous vein-images were captured using a modified web-camera that works well under the illumination of near-infrared light and a DSLR (Digital Single Lens Reflex) camera with an external filter to block visible light. An algorithm was developed to extract vein-segments from these images. The system will be used for automated delivery of haemostatic agent.