AI Algorithm Detects COVID-19 from Chest X-rays with 98% Accuracy
Researchers have developed an AI algorithm that can accurately detect COVID-19 infection from chest X-rays with an impressive 98% accuracy rate. This breakthrough could significantly improve the diagnostic process by quickly distinguishing between normal X-rays and those from individuals with pneumonia, which often displays similar symptoms to COVID-19.
Real-time reverse transcription-polymerase chain reaction (RT-PCR) testing is currently the most common method used for diagnosing COVID-19 infection. However, it has its limitations, including high costs, slow results, and the possibility of false negatives. To supplement this testing method, CT scans and chest X-rays play an important role in timely detection and management of contagious infections, especially when RT-PCR results come back negative.
COVID-19 demonstrates specific radiological indicators in chest X-rays, which radiologists utilize to diagnose viral infection. However, manually examining X-rays for signs of infection is time-consuming and prone to human error. To address these challenges, researchers at the University of Technology Sydney (UTS) harnessed the power of artificial intelligence (AI) to streamline the diagnostic process.
Amir Gandomi, the corresponding author of the study, explained, The most widely used COVID-19 test, real-time polymerase chain reaction (PCR), can be slow and costly and produces false negatives. To confirm a diagnosis, radiologists need to manually examine CT scans or X-rays, which can be time-consuming and prone to error.
Furthermore, the overlapping symptoms of COVID-19 – fever, cough, difficulty breathing, and sore throat – can make it challenging to differentiate the infection from other respiratory viral diseases like the flu or pneumonia.
In recent years, machine learning algorithms have gained popularity in the field of medicine. They have assisted doctors in diagnosing Parkinson’s disease, detecting breast cancer, and predicting stroke and heart failure. Deep learning, a subfield of AI, is especially effective in creating models that deliver accurate results based on input data without requiring manual feature extraction. In this study, researchers developed a deep-learning-based algorithm called a Custom Convolutional Neural Network (Custom-CNN) that specifically focuses on diagnosing COVID-19.
The researchers utilized two freely available chest X-ray datasets to train and test the AI model. These datasets consisted of three categories of chest X-ray images: normal, coronavirus-positive, and viral pneumonia. The Custom-CNN model was trained with 80% of the total images, while the remaining 20% were used for testing.
The objective of the study was to evaluate the model’s effectiveness in examining various relationships, such as coronavirus vs. viral pneumonia, normal vs. viral pneumonia, and coronavirus vs. normal. The results demonstrated that the Custom-CNN model achieved an impressive classification accuracy of 98.19% in distinguishing COVID-19, normal, and pneumonia X-ray samples. When compared to other models, the Custom-CNN outperformed them significantly.
Gandomi emphasized, Deep learning offers an end-to-end solution, eliminating the need to manually search for biomarkers. The Custom-CNN model streamlines the detection process, providing a faster and more accurate diagnosis of COVID-19.
Early diagnosis of COVID-19 infection is crucial to ensure patients receive the appropriate treatment, including antivirals, which are most effective when administered within five days of symptom onset. Additionally, early diagnosis encourages individuals to isolate themselves and protect others from potential infection.
Gandomi added, 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.
This groundbreaking development in AI technology has the potential to revolutionize the diagnosis of COVID-19 infections. The accurate and rapid detection provided by the AI algorithm could improve patient outcomes and assist healthcare professionals worldwide in navigating the challenges posed by the ongoing pandemic.
