Imaging is critical for assessing cancer treatment response; radiologists typically evaluate changes in the size of cancer lesions to determine if therapies are working. But with the advent of new treatments such as immunotherapy, better quantitative imaging indicators of treatment response are needed, according to senior author Dr. Daniel Rubin of Stanford University.
Many groups are creating algorithms to compute a variety of these new quantitative imaging biomarkers, but there are few tools to bring them into the routine workflow of radiologist image interpretation, Rubin said. As a result, the researchers sought to develop an image annotation and analysis plug-in for the freely available electronic Physician Annotation Device (ePAD) web-based quantitative imaging software platform. The plug-in computes imaging biomarker assessments on longitudinal imaging studies and then analyzes and compares treatment response based on those biomarkers.
They evaluated the ePAD plug-in on 8,199 annotations from 429 cancer patients who had one baseline CT scan and several follow-up studies with routine radiologist image assessments. The software automatically generated graphical treatment summaries helpful to oncologists based on the routine cancer lesion measurements they make on images, Rubin said. It also automatically computed new quantitative biomarkers without disrupting the routine workflow.
"ePAD permits automated summary of the response of patient cohorts to treatment as well as comparing response based on using different imaging biomarkers of response," Rubin told AuntMinnie.com. "It may, thus, give valuable insight into the effectiveness of different imaging biomarkers for use in clinical trials. The charts can also be exported easily to be used in radiology reports."
Take in this talk by presenter Emel Alkim to learn more.
