Artificial Intelligence

Google’s AI research group has shown that deep-learning algorithms can fine-tune ophthalmologists’ diagnosis of diabetic retinopathy on retinal fundus photographs, according to a study slated for publication in Ophthalmology. In the study, the physicians using the algorithm bested both AI alone and unassisted physicians on accuracy.

When manually corrected by radiologists, an AI system for automatically detecting and segmenting colorectal metastases in the liver can improve interpretative efficiency, according to a study published online March 13 in Radiology: Artificial Intelligence.

Radiology patients are confident artificial intelligence will improve healthcare workflow and efficiency, but they’re skeptical of the tech itself and remain unsure of how AI will factor into the patient experience, according to a study published online March 14 in the Journal of the American College of Radiology.

Authors of the research, published in the Korean Journal of Radiology, analyzed 516 published studies and found only six percent (31 studies) externally validated their AI.

An AI approach developed by Dutch researchers performed similarly to radiologists at detecting breast cancer, according to a multi-center, multi-dataset study published March 5 in the Journal of the National Cancer Institute.

The researchers believe their algorithm could help pathologists classify the histologic patterns of lung adenocarcinoma—the most common form of the disease—and potentially lead to more accurate staging.

The American College of Radiology (ACR) is asking for comments to be submitted by April 15.

Google and its sister company Verily announced on Monday, Feb. 25, the development of an AI-based algorithm that can screen eye imaging exams for diabetic retinopathy and diabetic macular edema—two of the leading causes of preventable blindness in adults with diabetes, according to a recent report by CNBC.  

A new AI-based software called the Ensemble Algorithm with Multiple Parcellations for Schizophrenia Prediction, or EMPaSchiz, can identify schizophrenia on fMRI scans with 87 percent accuracy, according to a recent report by AI in Healthcare.  

Mammography is an essential screening and diagnostic tool for the detection of breast cancer and the assessment of breast density. But, according to Victoria L. Mango, MD, a breast radiologist at Memorial Sloan Kettering Cancer Center in New York City, AI can help breast imagers and physicians see beyond basic breast density information provided by mammographic images and improve clinical management overall.

Researchers have created a machine learning model that identified 98 percent of malignant atypical ductal hyperplasia (ADH) lesions prior to surgery, according to a single-center study published in JCO Clinical Cancer Informatics. The approach saved 16 percent of women from unnecessary surgery.

There is an "immaturity" between machines and humans, said Paul J. Chang, MD, of the University of Chicago. Unless radiology departments augment their current IT infrastructure, AI could become another technological-driver of burnout.