A new study involving an AI radiology tool developed by Northwestern Medicine has been shown to enhance productivity and detect life-threatening conditions swiftly, offering a solution to the worldwide shortage of radiologists.
During a five-month period last year, the AI tool was deployed across Northwestern Medicine’s 11-hospital network, analysing around 24,000 radiology reports.
Go deeper with GlobalData
The research team, led by Dr Mozziyar Etemadi, assistant professor of anaesthesiology at Northwestern University Feinberg School of Medicine and of biomedical engineering at Northwestern’s McCormick School of Engineering, compared the radiograph report creation time as well as accuracy with and without using the AI system.
The study revealed an average increase of 15.5% in radiograph report completion efficiency, with few radiologists experiencing gains up to 40%, without any loss of accuracy.
Subsequent research indicates efficiency improvements of up to 80% and applicability to computed tomography (CT) scans.
This time-saving aspect is particularly crucial in urgent cases where prompt diagnosis is vital.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataThe authors of the study highlight that this is the world’s first generative AI radiology tool integrated into the clinical workflow.
It claims to demonstrate high accuracy across various X-ray types.
The university noted that the holistic model analyses complete X-ray or CT scans, generating near-complete reports in the radiologist’s style for final review and completion.
The AI system not only improves efficiency but also identifies critical conditions like pneumothorax [collapsed lung] in real-time.
It monitors reports for critical findings and cross-references patient records, alerting radiologists to urgent cases.
The team is further refining the model to identify delayed or overlooked diagnoses such as early-stage lung cancer.
The university noted that the AI tool is designed to assist radiologists in clearing backlogs and delivering quick results, but it is not intended to replace human expertise.
With two patents approved and others pending, Northwestern’s technology is in the initial stages of commercialisation.
Dr Etemadi said: “This is, to my knowledge, the first use of AI that demonstrably improves productivity, especially in health care. Even in other fields, I haven’t seen anything close to a 40% boost.”
