AI helps spot fatty liver disease on routine radiographs
With the help of artificial intelligence, chest radiographs could present a simple, affordable and noninvasive way to identify fatty liver disease.
Hepatic steatosis, or fatty liver disease, often does not present symptoms in its early stages and is sometimes detected incidentally when patients undergo CT imaging for other clinical concerns. If left undiagnosed and untreated for a prolonged period, it can progress to more serious conditions, such as cirrhosis and liver cancer, making its early detection crucial for positive outcomes.
Fatty liver disease is typically detected via CT, ultrasound or MRI. Though the modalities are reliable, they are not as easily accessible or affordable as standard X-ray imaging. Researchers from the Osaka Metropolitan University in Japan believe they have found a way to spot signs of fatty liver disease on routine chest X-rays, which include the top portion of the liver. They shared their findings in Radiology: Cardiothoracic Imaging, revealing their model achieved promising results.
For the study, the team retrospectively analyzed the cases of more than 4,400 patients who underwent both chest X-rays and controlled attenuation parameter (CAP) examinations at two institutions from November 2013 to May 2023. CAP values were used to determine whether patients had hepatic steatosis. A deep learning-based model was trained, tested and validated using patients’ imaging and medical data.
Model testing included internal and external datasets, and it performed well in both cohorts. The AUC, accuracy, sensitivity, and specificity for the internal test set were 0.83, 77%, 68%, and 82%; for the external test set, these figures were 0.82, 76%, 76%, and 76%, respectively.
Though the model could undergo some fine-tuning to refine its accuracy, the team is optimistic about their findings, suggesting the model has the potential to improve the early diagnosis of fatty liver disease.
“The development of diagnostic methods using easily obtainable and inexpensive chest X-rays has the potential to improve fatty liver detection,” research leader Sawako Uchida-Kobayashi, MD, an associate professor at Osaka Metropolitan University’s Graduate School of Medicine, and colleagues wrote. “We hope it can be put into practical use in the future.”
Learn more about the model here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.