Researchers at the University of Technology Sydney (UTS) Data Science Institute have developed an artificial intelligence (AI) system that can rapidly detect COVID-19 from chest X-rays with more than 98% accuracy. The study, recently published in Scientific Reports, highlights the pressing need for effective automated tools to detect COVID-19, given its significant impact on public health and the global economy.

The current gold standard for COVID-19 testing, real-time polymerase chain reaction (PCR), is slow, costly, and prone to producing false negatives. To confirm a diagnosis, radiologists manually examine CT scans or X-rays, which can be time-consuming and subject to human error. The newly developed AI system aims to address these challenges and could be particularly beneficial in countries experiencing a high number of COVID-19 cases where there is a shortage of radiologists. Chest X-rays are portable, widely available, and expose patients to lower levels of ionizing radiation compared to CT scans.

Common symptoms of COVID-19, such as fever, cough, difficulty breathing, and sore throat, can often be difficult to distinguish from flu and other types of pneumonia. The AI system uses a Custom Convolutional Neural Network (Custom-CNN), which is a deep learning-based algorithm. This Custom-CNN model can quickly and accurately differentiate between COVID-19 cases, normal cases, and pneumonia in X-ray images. Deep learning technology offers an end-to-end solution, eliminating the need for manual searching of biomarkers. The Custom-CNN model streamlines the detection process, providing a faster and more accurate diagnosis of COVID-19.

In cases where a PCR or rapid antigen test displays a negative or inconclusive result due to low sensitivity, patients may require further examination through radiological imaging to confirm or rule out the presence of the virus. The new AI system could prove beneficial in these situations.

While radiologists play a critical role in medical diagnosis, AI technology can assist them in making accurate and efficient diagnoses, as mentioned by Professor Amir H Gandomi, the corresponding author of the study. The performance of the Custom-CNN model was evaluated through a comprehensive comparative analysis, with accuracy as the criterion for assessment. The results demonstrated that the new model outperforms other AI diagnostic models.

A fast and accurate diagnosis of COVID-19 is crucial in ensuring that patients receive the appropriate treatment, including COVID-19 antivirals, which are most effective when taken within five days of symptom onset. Additionally, accurate diagnosis can help patients isolate themselves and prevent the further spread of the virus, ultimately reducing the frequency of pandemic outbreaks.

This breakthrough represents a significant step in addressing the challenges posed by the ongoing COVID-19 pandemic. The newly developed AI system has the potential to transform the landscape of COVID-19 diagnosis and management, aiding in the prompt and accurate detection of the virus.