Can CT-Based AI Provide an Objective Biomarker of Chronic Stress?

While chronic stress has been identified as a contributing factor for conditions ranging from heart disease and depression to obesity, there is no validated objective biomarker for providing a cumulative measurement of chronic stress, noted Shadpour Demehri, M.D., in a recent interview with Diagnostic Imaging.

“As you look throughout the whole spectrum and gamut of medicine, a lot of chronic diseases, especially in the older adult (and) middle-aged adults, have been linked to chronic stress. The evidence is very shaky, and the publications are not really reliable in a sense that the element of chronic stress has always been evaluated subjectively,” pointed out Dr. Demehri, a professor of radiology at the Johns Hopkins University School of Medicine.

However, in a recent study to be presented at the Radiological Society of North America (RSNA) annual meeting, Dr. Demehri and colleagues examined the potential association between deep learning assessment of the adrenal volume index (AVI) based off chest computed tomography (CT) scans for 2,842 study participants.

The researchers found that each 1 cm³/m² increase in AVI was linked to over four percent higher risks for heart failure and mortality.

“We were able to show that adrenal volume was associated with patient mortality and heart failure over the course of about 15 years of follow up,” noted lead study author Elena Ghotbi, M.D., a postdoctoral research fellow at the Johns Hopkins University School of Medicine.

(Editor’s note: For additional coverage of the RSNA conference, click here.)

Dr. Demehri is excited about the potential of CT-based AI assessment of AVI as an objective biomarker of chronic stress and emphasizes the ease of obtaining this data from standard chest CTs.

“ … We extract this (data) from the chest CT that's being done in a majority of middle-aged adults or older adults for various screening purposes, like cardiac calcium score (and) lung cancer screening. And this is something that you could use at zero additional cost or radiation exposure to come up with a measure of their body exposure to chronic stress, and that's objective. That's not based on a questionnaire, and that's not also subject to variability as the cortisol level is,” maintained Dr. Demehri.

(Editor’s note: For related content, see “Multimodal AI with CCTA and MRI Data Shows Promise in Predicting MACE in Patients with Obstructive CAD,” “Study Shows Merits of CTA-Derived Quantitative Flow Ratio in Predicting MACE” and “Combining Advances in Computed Tomography Angiography with AI to Enhance Preventive Care.”)

For more insights from Drs. Demehri and Ghotbi, watch the video below.