Mirko Zimic of Universidad Peruana Cayetano Heredia in Peru will develop a simple, low-cost lung ultrasound device that can automatically diagnose bacterial pneumonia and measure nutritional status in children in low-resource settings. They have developed an algorithm to detect lung infiltrates from digital ultrasound images as evidence of pneumonia in children, and have built a prototype device, which comprises an ultrasound probe connected to a laptop or smartphone for realtime analysis of the images. Visible and audio alerts are used to notify the technician of suspect regions. They will develop an additional algorithm to automatically measure anatomical parameters including rib diameter and skin thickness for predicting levels of malnutrition, and test it in a hospital in Peru. Ultimately, this will be incorporated into the ultrasound device so that both pneumonia and malnutrition can be easily diagnosed together by non-specialized health care workers.
More information about New Ways to Reduce Childhood Pneumonia Deaths Through Delivery of Timely Effective Treatment (Round 15)