Abstract:
Objective To explore the consistency and influencing factors between diffusion-weighted imaging (DWI) and magnetic resonance angiography (MRA) in diagnosing acute cerebral infarct (ACI).
Methods Ninety-eight suspected ACI patients who received treatment at Lai'an Jianing Hospital from January 2020 to February 2022 were selected as the study subjects and included in the training set. Among them, 58 were males and 40 were females, aged 45–80 (60.5±3.3) years old. Suspected ACI patients (33 cases) diagnosed and treated at Lai'an Jianing Hospital from March to October 2022 were selected for retrospective analysis according to the same criteria and included in the validation set. Among them were 18 males and 15 females, aged 42–79 (61.1±3.6) years old. Using clinical comprehensive diagnosis as the "gold standard", we analyzed the diagnostic efficacy, imaging manifestations, and consistency of examination results of DWI and MRA. Two independent sample t-tests were used for intergroup comparison of econometric data. The intergroup comparison of counting data was conducted using χ2 test. Kappa test with multiple classification data was performed to analyze the consistency between DWI and MRA in diagnosing ACI. Multiple Logistic regression analysis was conducted to screen for independent risk factors with inconsistent results between DWI and MRA examinations. Empower Stats and statistical software package "R" were used to draw a forest map, construct a risk column-chart prediction model, and evaluate the model. The discriminability and calibration of the risk-prediction model were determined using the receiver operating characteristic (ROC) curve and the Hosmer–Lemeshow goodness-of-fit test. Risk nomogram prediction model accuracy was evaluated using decision curve analysis.
Results Among the 74 patients diagnosed with ACI clinically, 73 (98.65%) were positive for DWI and 71 (95.95%) were positive for MRA. The difference in apparent diffusion coefficient (ADC) values between the healthy (≤6 h: (1.06±0.24)×10–4 cm2/s; 6–24 h: (1.13±0.26)×10–4 cm2/s; 24–72 h: (1.05±0.17)×10–4 cm2/s) and affected (≤6 h: (0.59±0.11)×10–4 cm2/s; 6–24 h: (0.44±0.10)×10–4 cm2/s; 24–72 h: (0.53±0.09)×10–4 cm2/s) brain tissues of patients were statistically significant (t=10.227, 12.630, 7.646; all P<0.05). Within 24 h after the onset of the disease, the ADC value and rADC (≤6 h: (0.53±0.08); 6–24 h: (0.43±0.05)) in the affected side of the brain initially decreased significantly (t=5.410, 5.569; both P<0.05) and then increased significantly (24–72 h ADC: (0.53±0.09)×10–4 cm2/s, 24–72 h rADC: (0.49±0.06)) (t=2.274, 2.835; both P<0.05). A total of 68 patients had consistent results between DWI and MRA (Group A), whereas 30 had inconsistent ones (Group B). The consistency between DWI and MRA was good (Kappa=0.654, P<0.05). Results of multivariate Logistic regression analysis showed that onset time ≤24 h, posterior circulation, length of infarct lesion <2 cm were independent risk factors for inconsistent results between the DWI and MRA diagnosis of ACI patients (OR=1.119, 1.169, 1.567; all P<0.05). Evaluation results of the risk nomogram prediction model showed that its discrimination, accuracy, and effectiveness were all high, and the area under curve of the training and validation sets were 0.930 (95%CI: 0.899–0.961, P<0.001) and 0.855 (95% CI: 0.812–0.898, P<0.001).
Conclusions DWI can clearly display the location and degree of ischemia of the lesion, whereas MRA can accurately locate the infarcted blood vessels and their stenosis. The consistency between the two examinations is good, and both can help diagnose and evaluate ACI. The onset time, posterior circulation, and length of infarct lesion are risk factors that affect the consistency of diagnosis between the two.