By combining AI, high -end microscopy and engineering, the team revolutionizes how the infection is diagnosed, antibiotic prescription methods.
challenge
Antibiotics are less effective. Excessive use of wide spectrum drugs and often delayed diagnosis, which affects antimicrobial resistance, and is already in charge of more than 1.2 million people every year. However, a wide range of drugs are continuously prescribed because the current test to identify bacteria and antibiotic sensitivity takes more than 24 hours, delayed treatment and worsening results. Therefore, finding a test that can quickly identify the pathogen and sensitivity means that the patient can be treated much faster with a narrow spectral antibiotic aimed at a specific bacteria.
access
The Oxford Martin program for antimicrobial resistance tests develops a quick diagnostic test that collects researchers in physical, engineering and medical science for 30 minutes.
Infection tests train the online users to train the AI model to help you learn what antibiotics are and what antibiotic resistant bacteria are in the microscope.
The team also started an infection test, a global citizen science project to train the AI model, and was classified as more than 5,000 volunteers using more than 1 million images.
influence
- 30 minutes results: To identify effective antibiotics with bacterial species, it is 50 times faster than the current standard.
- AI -centric accuracy: A trained deep learning model -75% or more accuracy is trained to distinguish whether bacteria are sensitive to antibiotics based on visual changes.
- Portable and Accessibility: Designed to be used in a field tent that provides services to the hospital, pharmacy or victims of natural disasters.
- Target treatment. It enables accurate and personalized use of narrow spectrum antibiotics.
What is the following
This test is already working in laboratory settings and is being stylish for speed, offline function and wider antibiotic application. The next step is to develop commercial scale with global distribution targets.
Professor Achilles Kapanidis, who led the work with his fellow Chris Nellåker, Nicole Stoesser, Monique Andersson and Derrick Crook, explains:
“The average time for bacteria to develop resistance to new antibiotics is about 2 years, so it is a very precious resource. In sub -Saharan -south Africa, Southeast Asia, and South America, the effects of antibacterial resistance are already very high, and if death can reach $ 1 million a year by 2050, this new test can actually expand. To protect the future, it is necessary to prevent the use of extensive antibiotics. ”
