Diffusion kurtosis imaging, paired with a radiomics model, could reduce unnecessary breast biopsies

Diffusion kurtosis imaging, a new MRI breast imaging technique that does not require a contrast agent, and a radiomics breast cancer model could help reduce false-positive findings and unnecessary biopsies, according to a new study published in Radiology.

The researchers, including co-lead authors Sebastian Bickelhaupt, MD, of the German Cancer Research Center, and Paul Ferdinand Jaeger, MSc, studied 222 consecutive patients examined at two different facilities. All patients underwent breast screening mammography and presented with a suspicious lesion (BI-RADS category 4 or 5) and an indication for breast biopsy. The women underwent kurtosis diffusion-weighted imaging (DWI), and a radiomics breast cancer model was developed with that data.

Bickelhaupt explained the importance of trying diffusion kurtosis imaging in a prepared statement distributed by RSNA. “Diffusion kurtosis imaging has been introduced in DWI to provide important information on tissue structures at a microscopic level,” he said. “Since malignant lesions disrupt the tissue structures at this level, diffusion kurtosis might serve as a relevant marker of changes.”

So, was it successful? Overall, that radiomics model was found to reduce false-positive results by 70 percent while retaining a sensitivity of more than 98 percent.

“The model might help to lower the number of BI-RADS 4 lesions suspected of being cancer on the basis of screening mammography while retaining a high sensitivity similar to the sensitivity reported for biopsies themselves,” Jaeger said in the same prepared statement.

BI-RADS 4 lesions were found to benefit significantly from using the model, but not BI-RADS 5 lesions.

“Analysis of the radiomics model shows that all clarified false-positive results were BI-RADS 4a or 4b and the single benign lesion categorized as BI-RADS 5 was not identified correctly,” the authors wrote.