Diagnostic Utility of Non-contrast Thoracic Computed Tomography in the Detection of Anemia and Its Severity: A...

Background Anemia is a widespread clinical condition that detrimentally influences both quality of life and survival outcomes, particularly in elderly and chronically ill patients. Recent evidence highlights the potential of unenhanced thoracic computed tomography (CT) scans as a non-invasive modality for detecting anemia by assessing blood attenuation values. Our study aims to determine the diagnostic accuracy of unenhanced thoracic CT scans in predicting anemia. Methods This was a prospective, single-center, hospital-based analytical cross-sectional study conducted in the Department of Radiology at Meenakshi Medical College Hospital and Research Institute, Kanchipuram District, Tamil Nadu, India, between June 2024 and December 2024. Results Among the 225 patients included in the study, 125 (55.6%) were found to have anemia, classified as mild (22.7%), moderate (25.3%), or severe (7.6%). Anemic individuals were significantly older than their non-anemic counterparts (mean age: 59.3 vs. 52.5 years; p = 0.005), though gender distribution showed no significant difference (p = 0.179). Blood attenuation measured in Hounsfield units (HU) on unenhanced thoracic CT was significantly reduced in anemic patients across all vascular sites assessed, including the aortic arch (+33.4 vs. +45.2 HU), ascending aorta (+35.2 vs. +45.1 HU), descending aorta (+36.4 vs. +44.2 HU), and pulmonary artery (+36.3 vs. +45.0 HU), all with p-values < 0.001. Subjective CT signs were also more frequent in the anemic group, with the aortic ring sign present in 42.4% compared to 11.0% in non-anemic patients, and the interventricular septum sign observed exclusively among anemic individuals (16.8%). Hemoglobin levels correlated strongly with HU values across vascular sites, most notably at the aortic arch (r = 0.887), with similarly high correlations in the ascending aorta (r = 0.835) and pulmonary artery (r = 0.787). Receiver operating characteristic (ROC) analysis confirmed excellent diagnostic accuracy, with areas under the curve (AUCs) of 0.953 for anemia at the aortic arch and 0.944 for severe anemia at the pulmonary artery, supporting thoracic CT as a valuable non-invasive tool for anemia detection.  Conclusion Unenhanced thoracic CT scans demonstrate high diagnostic accuracy in predicting anemia through both objective blood attenuation values and subjective imaging signs. These findings support their use as a rapid, non-invasive adjunct in clinical anemia assessment.