AI model can help alleviate increased sonographer workload, but still has high missed diagnosis rate.
Ultrasound – augmented with a new artificial intelligence (AI)-based computer-aided detection model – can be used to accurately predict breast cancer cases, according to recently published research.
In a study published on June 9 in the Journal of Digital Imaging, a research team, from Sichuan University’s West China Hospital, revealed that even though the AI model has a high missed diagnosis rate, it performs at roughly the same level as trained sonographers. Consequently, it could be used to alleviate provider workload – but not substitute for the provider altogether.
“Although the number of sonographers who perform ultrasonographic examinations, interpret the images, and issue diagnostic reports has increased,” wrote the team, led by Heqing Zhang from the ultrasound department in West China Hospital, “currently, they cannot keep up with the growth in the requirement of ultrasound examinations.”
This ever-growing volume of case studies opens the door for sonographers to make more mistakes. AI models could help solve this problem, the team proposed.
To test the feasibility of using AI models, the investigators used a training set of 5,000 breast ultrasound images – half which were malignant – to develop four convolutional neural networks. They used just over 1,000 scans to test their prediction model.
According to the study results, one AI model – the Inception V3 – outperformed the sonographers with a better area under the curve score, but it was not as sensitive. The providers had a sensitivity rate of 90 percent and missed fewer than 10 percent of diagnoses.
Further work is needed to improve the model’s sensitivity, the team said, because missed breast cancer diagnoses can lead to more negative outcomes than a misdiagnosis.
Current Insights and Emerging Roles for Contrast-Enhanced Mammography
May 10th 2024In a recent lecture at the 2024 ARRS Annual Meeting, Jordana Phillips, MD, discussed the role of contrast-enhanced mammography in staging breast cancer, evaluating response to neoadjuvant chemotherapy and recalls from screening.
Study Finds High Concordance Between AI and Radiologists for Cervical Spine Fractures on CT
May 6th 2024Researchers found a 98.3 percent concordance between attending radiology reports and AI assessments for possible cervical spine fractures on CT, according to new research presented at the 2024 ARRS Annual Meeting.
The Reading Room: Artificial Intelligence: What RSNA 2020 Offered, and What 2021 Could Bring
December 5th 2020Nina Kottler, M.D., chief medical officer of AI at Radiology Partners, discusses, during RSNA 2020, what new developments the annual meeting provided about these technologies, sessions to access, and what to expect in the coming year.
New Literature Review Finds ChatGPT Effective in Radiology in 84 Percent of Studies
April 29th 2024While noting a variety of pitfalls with the chatbot ranging from hallucinations to improper citations, the review authors found the use of ChatGPT in radiology demonstrated “high performance” in 37 out of 44 studies.