Radiomics an excellent predictor of whether lung nodules are cancer or not

May 17, 2018 | Publications

A new article published in PLOS|One shows the power of Radiomics as a predictor. In this case the researchers were able to show that a radiomic signature made up of eight different classifiers had the ability at a very high sensitivity (0.904) and specificity (0.855) to classify lung nodules as malignant or non-malignant. Many healthcare systems are employing cancer screening programs that are producing a large number of incidentally found nodules/lesions, and being able to determine a diagnosis of these without painful, dangerous, and expensive procedures could save patients a lot of suffering and anxiety and greatly reduce costs for the healthcare system.

The full text article is available here.

The HealthMyne Platform was designed from the ground up to provide clinical decision support (CDS) when radiomic signatures like these are discovered. As soon as a lesion is identified at the point-of-read, RPM functionality determines the full radiomic profile and can match that to a signature within seconds and then automatically include that information in the radiology report. This is only one of the enhanced CDS workflows available.