Artificial Intelligence

Combat antibiotic-resistant bacteria with artificial intelligence

Combat antibiotic-resistant bacteria with artificial intelligence

An artificial intelligence system applied to medicine could analyze large volumes of data effectively, allowing human decision making in the short term, and obtaining the right medicine to counteract the disease under study.

Among the most encountered medical problems is the presence of bacteria resistant to antibiotics, which, in the absence of an effective solution, there are patients who die when they contract an infectious condition of this kind.

Although we currently have resources to combat antibiotic resistant bacteria, it is the waiting time while finding the right and effective antibiotic, which could complicate a medical condition, not counting the time it takes the laboratory to cultivate and analyze samples, along with drug tests that are performed to achieve effectiveness, this whole process is expensive, and some patients cannot wait because of the critical condition that may present.

With fast and effective results in mind, a group of scientists at ETZ Zurich University in Switzerland trained artificial intelligence system algorithms with mass spectrometry data to teach them the automatic detection of antibiotic-resistant bacteria.

During the research, the scientists sought to obtain results quickly, even up to 24 hours earlier compared to traditional tools, allowing physicians to begin more accurate and effective treatment of their patients.

The system was trained using a database of more than 300,000 mass spectra of individual bacteria, covering around 800 different bacteria and more than 400 different antibiotics, enabling the algorithm to detect resistance automatically. The algorithm automatically detects the resistance, solving important and necessary questions when working quickly with a patient who presents a serious clinical picture in terms of infections.

For now, the research has been conducted in laboratories, the team plans to present a clinical trial with results in real patients.

More information at https://ethz.ch/en/news-and-events/eth-news/news/2022/01/ai-offers-a-faster-way-to-predict-antibiotic-resistance.html

08 de Febrero, 2022



metodika