While medical imaging has long been an important tool in the diagnosis of many diseases, the COVID-19 pandemic has set developers scrambling to develop new methods of quickly detecting the infection.
To that end, researchers at the Los Angeles-based Terasaki Institute for Biomedical Innovation (TIBI) researchers have used AI to develop and validate an image-based detection model for COVID-19.
In recent years, AI models have been increasingly incorporated into imaging technology to improve diagnostic capabilities by detecting disease characteristics that are often not visible to the naked eye.
In a multi-institute collaborative effort, TIBI researchers led a multi-institute collaborative effort to develop an AI image-based detection model, first using a model to automatically collect imaging data from the lung lobes, then analyzing the data to identify features as potential diagnostic biomarkers for COVID-19.
The diagnostic biomarkers enabled researchers to differentiate COVID-19 patients from pneumonia and healthy patients. The model was developed using a cohort of 704 chest x-rays and validated with 1597 cases from multiple sources made up of healthy, pneumonia, and COVID-19 patients. The results indicated the model was successful in classifying diagnoses of various patients.
“This highly advanced artificial intelligence model further helps our ability to precisely detect COVID-19 patients. In addition, such a model can be applied for diagnosis of other diseases using different imaging modalities,” lead researcher Samad Ahadian, PhD, said in a statement.
According to the researchers, using computer modeling with data extracted from medical images shows promise in improving precision medicine and could revolutionize medical practice in clinics. Moreover, creating methodologies to capture full sets of information while suppressing irrelevant features will enhance the reliability of AI models.
“Artificial intelligence-driven models with diagnostic and predictive capabilities are a powerful tool that are an important part of our research platforms here at the institute,” said TIBI CEO and Director Ali Khademhosseini, PhD, Director and CEO of TIBI. “This will carry over into countless applications in the biomedical field and in the clinic.”
Not surprisingly, given the ongoing pandemic, there has been a high demand for rapid and accurate methods of COVID-19 infection detection. Currently, the primary method is to use reverse transcription-polymerase chain reaction (RT-PCR) on samples collected from nasal or throat swabs, but this method can lead to inaccuracies due to sampling errors, low viral load, and sensitivity limitations. Inaccuracies are especially significant for patients in the early stages of infection.
Photo by Devrimb/Getty Images