Revolutionary AI detects COVID-19 quickly and accurately

A revolutionary Artificial Intelligence (AI) system has been developed by researchers, capable of swiftly and accurately detecting COVID-19 from chest X-rays with an accuracy exceeding 98%. Recently published in Nature Scientific Reports, the study, spearheaded by Professor Amir H Gandomi from the Data Science Institute at UTS, addresses the critical need for automated tools in combating the global impact of COVID-19 on public health and the economy. 

Traditional COVID-19 testing methods, such as real-time polymerase chain reaction (PCR), can be slow, expensive, and prone to false negatives. To confirm diagnoses, manual examinations of CT scans or X-rays by radiologists are often required, introducing time-consuming processes vulnerable to errors. The new AI system, employing a deep learning-based algorithm called Custom Convolutional Neural Network (Custom-CNN), stands out by swiftly and accurately distinguishing COVID-19 cases, normal cases, and pneumonia in X-ray images. 

The new AI system could be particularly beneficial in countries experiencing high levels of COVID-19 where there is a shortage of radiologists. Chest X-rays are portable, widely available and provide lower exposure to ionizing radiation than CT scans

Professor Gandomi 

By offering an end-to-end solution through deep learning, the Custom-CNN model eliminates the manual search for biomarkers, streamlining the detection process and facilitating a quicker and more precise COVID-19 diagnosis. In situations where PCR or rapid antigen tests yield inconclusive results, the AI system could play a crucial role in further examinations through radiological imaging. 

The breakthrough, by the AI assist, holds promise for ensuring prompt and accurate diagnoses of COVID-19, enabling timely treatment, isolation, and contact tracing to mitigate the spread of the virus.