Quantitative Attenuation-Based Temporal Staging of Ischemic Cerebral Infarction on CT with MRI Correlation

Abstract

Background: Accurate temporal staging of ischemic cerebral infarction is essential for appropriate therapeutic decision-making and characterisation in acute stroke management. Non-contrast computed tomography (CT) remains the first-line imaging modality due to its rapid acquisition and widespread availability. However, staging is frequently based on subjective visual interpretation. Quantitative assessment using Hounsfield Unit (HU) attenuation values may provide an objective method for differentiating infarct stages. Given the increasing emphasis on quantitative imaging biomarkers in neuroradiology, objective attenuation-based stratification may reduce observer dependency and enhance reproducibility in routine stroke assessment.

Aims and Objectives: To evaluate the role of quantitative CT attenuation values in categorizing ischemic cerebral infarction into acute, subacute, and chronic stages with clinical correlation, and to derive stage-specific reference HU ranges for prospective CT-based categorization.

Methodology: A retrospective study was conducted on 254 patients diagnosed with ischemic cerebral infarction who underwent non-contrast CT over an 18-month period in a tertiary care hospital. Infarcts were classified into acute (6–48 hours), subacute (48 hours–14 days), and chronic (>14 days) stages based on clinical history and duration of symptoms. Three standardized regions of interest were placed within the infarcted parenchyma, and the mean attenuation value was calculated for analysis. Intergroup comparison was performed using appropriate statistical tests. Interobserver agreement was assessed in 30 randomly selected cases using intraclass correlation coefficient. A subset of 32 patients who also underwent subsequent magnetic resonance imaging was included for validation, and CT-based temporal staging was compared with MRI features of infarct evolution to assess concordance.

Results: Among 254 cases, 71 (28%) were acute, 102 (40%) subacute, and 81 (32%) chronic infarcts. The mean attenuation values of the ischemic region were 25.98 HU in acute, 18.76 HU in subacute, and 11.16 HU in chronic infarcts. The mean attenuation values were 25.98 ± 2.8 HU in acute, 18.76 ± 3.9 HU in subacute, and 11.16 ± 2.2 HU in chronic infarcts. Percentile-based distribution analysis demonstrated stage-specific attenuation intervals: acute infarcts (23.7–31 HU), subacute infarcts (14.5–24.5 HU), and chronic infarcts (8.3–14 HU). These attenuation distributions showed statistically significant differences among the three stages (p < 0.001). The derived percentile-based intervals provided clear separation across stages, supporting objective temporal stratification. Interobserver reliability was excellent, with an intraclass correlation coefficient of 0.91. In the MRI validation subset, CT-based staging demonstrated concordance with MRI findings in 29 of 32 cases (90.6%).

Conclusion: Quantitative Hounsfield Unit analysis on non-contrast CT provides an objective and reproducible method for temporal staging of ischemic cerebral infarction. The percentile-derived attenuation intervals demonstrated clear stage-wise separation and may serve as cohort-based reference standards for prospective CT-based categorization. This quantitative approach enhances diagnostic confidence in acute stroke evaluation and offers a practical imaging biomarker framework, particularly in settings where advanced imaging modalities are unavailable or contraindicated. The high concordance observed in the MRI validation subset further reinforces the biological reliability of attenuation-based temporal stratification and supports its potential application in routine clinical practice.