Inflammatory ratios and CT Hounsfield units as predictors of infarct size and severity in acute ischemic stroke

Background and objectives. Systemic inflammatory indices (neutrophil–lymphocyte ratio (NLR), platelet–lymphocyte ratio (PLR), lymphocyte–monocyte ratio (LMR), and mean platelet volume (MPV)) are linked to brain injury after acute ischemic stroke (AIS). CT Hounsfield units (HU) reflect tissue hypoattenuation and correlate with infarcted tissue and clinical outcome. We evaluated whether these blood indices are associated with HU, infarct volume, and National Institutes of Health Stroke Scale (NIHSS) at admission. Materials and methods. We performed a retrospective cross-sectional study of consecutive AIS patients treated at Dr. Wahidin Sudirohusodo Hospital between January 2023 and December 2024. We recorded demographics, risk factors, blood counts, HU (three oval ROIs within the infarcted region; mean of three), infarct volume (ABC/2 method on non-contrast CT), and NIHSS. Associations were examined using Spearman’s rho (p < 0.05). Results. Among 126 patients (mean age 56.8 years; 59.5% male), median indices were: NLR 3.75, PLR 142.34 (41.41–616.60), LMR 2.73, MPV 9.4 fL; median HU 20; median infarct volume 3.46 cm³; median NIHSS 9. NLR correlated positively with HU (r = 0.229, p = 0.010), infarct volume (r = 0.607, p < 0.001), and NIHSS (r = 0.506, p < 0.001). PLR correlated with infarct volume (r = 0.330, p < 0.001) and NIHSS (r = 0.247, p = 0.005). LMR correlated inversely with HU (r = −0.249, p = 0.005), infarct volume (r = −0.552, p < 0.001), and NIHSS (r = −0.417, p < 0.001). MPV correlated positively with infarct volume (r = 0.187, p = 0.036) and NIHSS (r = 0.214, p = 0.016) but showed no significant correlation with HU. Conclusions. Higher NLR, PLR, and MPV and lower LMR are associated with larger infarcts and greater neurological deficit; NLR and LMR also relate to HU hypoattenuation. Findings support integrating routine blood indices with CT to refine early risk stratification in AIS, acknowledging the limitations of retrospective, single-center correlations.