Unmet Need
There are 7.8 million children diagnosed with pharyngitis each year, and another 4.8 M children with other associated complaints (respiratory symptoms or fever). These typically warrant a rapid strep test. The CDC estimates that there are 11,000 to 24,000 cases of strep throat each year in the United States, and that between 1,200 and 1,900 people die from this illness due to pneumonia and streptococcal toxic shock syndrome. Globally, the burden from group A strep infections is even greater.
While telehealth medicine continues to revolutionize patient care, diagnostic tests for strep throat remain unavailable in the home setting. Therefore, there is a strong need to diagnose strep throat remotely or via telehealth consultations. This test would expand patient access to treatment and improve patient outcomes by decreasing the complications that inevitably arise from untreated strep throat.
Technology Overview
Inventors at Johns Hopkins have conceptualized a machine learning algorithm that can identify and diagnose a streptococcus pharyngitis infection (aka strep throat). This algorithm uses video/image
analysis and deep learning methods, combined with clinical criteria for the prediction. With this technology patients with strep throat can receive a diagnostic test at home, by uploading a picture or video of their throat exam using their mobile device, and analyzed by the machine learning algorithm.
Stage of Development
Proof of concept. Need to develop the image database and the algorithm.
Further data acquisition is ongoing.
