Australian researchers say AI support prevents misdiagnoses, speeds decisions, and could transform frontline emergency care.
Artificial intelligence is helping emergency department doctors make faster and more accurate diagnoses, according to new research led by the Australian Institute of Health Innovation at Macquarie University.
The study found that AI-assisted interpretation of chest x-rays improved diagnostic accuracy by 5.9% and enhanced clinical decision-making by 3.2%.
The greatest benefits were seen among senior resident medical officers, whose diagnostic accuracy increased by 11.8%, equivalent to preventing one misdiagnosis for every nine patients.
The research, published in the BMJ Group’s Emergency Medicine Journal, evaluated the use of an AI tool capable of identifying up to 124 different findings in chest x-rays.
The study involved 200 emergency doctors who assessed 18 clinical scenarios both with and without AI support. Diagnoses and patient management were compared, with and without AI, using clinical vignettes representative of emergency department patient presentations.
Lead researcher Professor Farah Magrabi, the AIHI’s Professor of AI and Patient Safety, said AI assistance in interpreting chest x-rays could ensure critical diagnoses are not missed or delayed in the emergency department.
The research was conducted in a simulated environment using the Harrison.ai Chest X-ray solution. The authors said it was one of only a few studies in the world to both demonstrate the impact of AI assistance in decision-making and also show improvement in clinician performance.
The vignettes completed by participants were randomly selected from a bank of 49 based on the Australasian College for Emergency Medicine curriculum, covering typical and important conditions for which chest x-rays would be indicated in emergency medicine.
These included cardiovascular presentations, such as congestive cardiac failure, disorders of pericardium and aortic dissection, as well as respiratory presentations like pneumonia, aspiration, pneumothorax, pneumomediastinum, pleural effusions, lung lesion/mass and chronic obstructive pulmonary disease.
Additional areas included oncological and immunological presentations (e.g., sarcoidosis), trauma and orthopaedic cases involving fractures and dislocations, and procedural indications such as advanced airway management, tube thoracostomy, nasogastric tube and central venous access.
“Vignettes with normal x-rays include several gastrointestinal presentations, pulmonary embolism and viral infections,” the authors wrote.
“This comprehensive coverage ensured the vignettes reflected the breadth of conditions encountered in the ED.”
Co-author Professor Michael Dinh, founder and director of the Royal Prince Alfred Hospital Green Light Institute for Emergency Care, said chest x-rays were a core diagnostic tool in emergency medicine.
“More than half of all patients admitted to emergency departments are ordered a chest x-ray, but as formal radiology reporting can be delayed by hours, clinicians are sometimes under pressure to interpret results themselves in order to inform urgent clinical decisions. This study has shown that AI can support this decision-making,” he said.
“Having an AI supported interpretation of the chest x-ray is a game changer for clinicians – enabling them to make on-the-spot decisions about diagnosis and support, ensuring patients receive faster and more appropriate care.”
With emergency departments under increasing pressure, on-the-spot reporting could support clinicians to identify life threatening conditions requiring immediate care, such as such as when diagnosing a punctured lung or pneumonia.
It could also help determine when it was safe to refer patients for less urgent care, such as for lung nodules or a mild chest infection, the researchers noted.
Professor Magrabi noted that rural or remote hospitals with fewer senior emergency consultants could particularly benefit from the added support provided by an AI interpretation of x-ray results.
Dr Mark Phillips, Harrison.ai’s chief clinical officer, said the research findings demonstrated the potential benefits of integrating AI into the clinical review of chest x-rays in emergency departments.
“It’s encouraging to see independent evidence demonstrating the impact of our technology in supporting better, safer care for patients,” he said.
“We’re already working with emergency departments across Australia, Asia, and Europe, and this study reinforces the value of AI in supporting clinicians on the front line of care.”
The study was funded by the Digital Health Cooperative Research Centre. CEO Annette Schmiede said the research was a “powerful example of the Digital Health CRC model in action, bringing together industry, research, and health services to deliver real-world impact”.
“Our partnership with Harrison.ai has supported the development of this world-class AI tool,” she said.
“Harrison.ai is an Australian success story, a homegrown company translating cutting-edge digital health research into clinical practice.
“By supporting clinicians with rapid and accurate interpretation of chest x-rays, this technology demonstrates how AI can transform emergency medicine, improve patient outcomes, and help deliver safer, more efficient care.”
The authors concluded that improvements in diagnosis and patient management without meaningful increases in interpretation time suggested AI assistance could benefit clinical decisions involving chest x-ray interpretation.
“Further studies [in real-world clinical settings] are required to ascertain if such improvements translate to improved patient care,” they said.
