Editor's Note: Publication of AI Research in Radiology.
Radiology. 2018 Dec;289(3):579-580. doi: 10.1148/radiol.2018184021. Epub 2018 Nov 6.
Since the beginning of our discipline, research in radiology has been related to hardware devices. We have been largely concerned with the performance of various imaging modalities (eg, CT, MRI, PET) to detect disease (eg, cancer, cardiovascular disease, trauma). Our research results tell us the sensitivity and specificity of hardware to diagnose disease. Research hardware may be a new type of CT or MRI scanner. Even a new molecular imaging contrast agent operates like hardware: Researchers seek to determine the impact of the contrast agent plus the scanner on diagnostic performance. We expect hardware research to eventually translate to our own imaging practices. If PET detects more metastases than CT for a certain disease, we eventually translate such results into practice guidelines.