Objective To investigate the diagnostic efficacy of CT perfusion imaging (CTP) combined with MRI diffusion-weighted imaging (DWI) in different TOAST subtypes of acute cerebral infarction (ACI), and provide a basis for clinical imaging classification. Methods A retrospective analysis was conducted on the data of 80 patients with ACI admitted to The First Affiliated Hospital of Henan University of Science and Technology from January 2022 to April 2025. All patients underwent CTP and DWI examinations within 24 hours of onset. Clinical comprehensive diagnosis was used as a reference standard to evaluate the sensitivity, specificity, accuracy, Kappa consistency, and AUC value under the receiver operating characteristic (ROC) curve of CTP, DWI, and their combination in the classification of ACI, and compare them statistically. Results In individual diagnosis, the sensitivity of CTP in LAA and CE types was 88.57% and 83.33%, respectively, while the sensitivity of DWI in SVO type was 94.12%. The accuracy of combined diagnosis in LAA, CE, and SVO types is 95.00%, 96.25%, and 96.25%, respectively, which is significantly higher than the individual diagnosis of CTP and DWI. The AUC values of the combined mode all exceeded 0.95, and the Kappa value reached 0.842, which was better than CTP (0.726) and DWI (0.693), and the difference was statistically significant (P < 0.05). Conclusion CTP combined with DWI has higher accuracy and consistency in TOAST classification of ACI, and is superior to single imaging techniques. It has important clinical application value in early classification diagnosis, etiological mechanism inference, and guiding personalized treatment, and is recommended as a routine combination method for stroke imaging assessment.