This research aimed to analyze the diagnostic value of diffusion tensor imaging (DTI) based on tensor image registration algorithm for hypoxia-ischemic encephalopathy (HIE). In this study, 90 newborns diagnosed with HIE who were admitted to our hospital from October 30, 2016 to January 30, 2019 were selected as experimental group (EG) and were divided into mild group (EG 1), moderate group (EG 2), and severe group (EG 3). In addition, 45 normal newborns of the same period were selected as the control group (CG). All subjects underwent DTI, which was processed by image registration algorithm, and the regions of interest (ROIs) were manually selected, including anterior limb of internal capsule (ALIC), posterior limb of internal capsule (PLIC), frontal white matter (FWM), centrum semiovale, corpus callosum (CC), and lenticular nucleus. Fractional anisotropy (FA) values for each ROI were recorded. Receiver operating characteristic (ROC) curve was used to analyze the accuracy, sensitivity, specificity, and area under curve (AUC) of each ROI in diagnosing HIE. The results showed that the images processed by the registration algorithm were clearer than the original images, the signal-to-noise ratio (SNR) was increased, and the artifact was decreased, so that the lesions of newborns with HIE could be clearly observed. FA values in some ROIs in EG were significantly lower than those in CG (P < 0.05). FA values of all ROIs in the EG 3 were significantly lower than those in the EG 1 and the EG 2, with significant differences (P < 0.05). The FA values of the ALIC, the PLIC, and the centrum semiovale in the EG 2 were significantly lower than those in the EG 1, with significant differences (P < 0.05). The accuracy, sensitivity, specificity, and AUC area of the ALIC were higher, that is, 87.6%, 87.7%, 89.6%, and 0.879, respectively, indicating that with the aggravation of the condition of HIE, the morphological damage of the white matter fiber bundle became more serious and the FA value decreased. Therefore, DTI imaging is safe and feasible in the diagnosis of HIT lesions, especially with high accuracy, sensitivity, and specificity in the ALIC and FWM.
Read full abstract