The extensive use of antibiotics (diagnostic and therapeutic) and the expansion of invasive instruments led to an unforeseen problem: the spread of nosocomial (hospital-acquired) infections. These generate needless expenses, and also reduce the quality of life of the patients, prolonging their recovery and promoting the resistance of pathogens against the antibiotics. In 2009, healthcare costs caused by HAIs where about 40 billion dollars in the U.S. and it is increasing every year, causing serious damage and even death in the most serious cases. Recent studies revealed that at least 30% of HAI could be preventable . The pathogens are most often transferred via the hands of the medical staff, therefore compliance with the hospitals hand hygiene protocol is absolutely necessary.
To help medical staff in the validation of their hand washing technique, we developed a compact, mobile device which performs an objective evaluation of hand disinfection. The hardware consists of a metal box with matte black interior with ultra-violet lighting and a digital camera. Image processing and segmentation is performed on a regular notebook to determine the hand washing quality. After the hand washing procedure, the surgeons use a UV-reflective powder, showing bright under UV light only on sterile surfaces. When the surgeon insert its hands into the box, the camera placed on the top takes an image of the hand for evaluation.
I created the software that performs the segmentation and clustering automatically. First, the hand contour is determined from the intensity image, than two clusters are built using a threshold value, derived based on the average intensity of the region of interest. The clustering methodology is based on a fuzzy c mean algorithm. As a result, if the affected area is significant, the system warns the surgeon to wash their hands again. The main advantage of the device is the ability to obtain objective result on the quality of hand disinfection.