Adoption of Ultrasonic Imaging Diagnosis Under Machine Learning Algorithm in Treatment of Pregnancy-Induced Hypertension and Nursing

Abstract

In order to study the application of doppler ultrasound in monitoring the changes of important vascular hemodynamic indexes in pregnant women with Pregnancy-induced Hypertension (PIH) during pregnancy under the Omaha systematic nursing intervention (NI), in this study, 83 PIH pregnant women were taken as the study object. First, 60 normal pregnant women were taken as the control group (CG). Ultrasonic images of the uterine artery (UA) and middle cerebral artery (MCA) were collected. The machine learning (ML) algorithm was used to detect the edge of the target vessel in the UA image. The hemodynamic indexes of UA and MCA were compared between the two groups. Then, according to the routine NI and the NI method based on the Omaha system, 83 PIH pregnant women were divided into the routine group (RG) (n = 35) and the observation group (OG) (n = 48). The differences in UA hemodynamic parameters and quality of life (QOL) scores were compared between the two groups at 6 d after surgery. The results showed that pre-pregnancy BMI, mean artery pressure (MAP), and uric acid content in the CG were significantly lower than those in the PIH group (P < 0.05); the content of serum albumin in the CG was significantly higher than that in the PIH group (P < 0.05); compared with Sobel operator, edge detection (ED) based on ML algorithm was more coherent; in the CG, the systolic peak velocity (PSV), acceleration index (AI), pulse index (PI), and resistance index (RI) were significantly lower than those in the PIH group (P < 0.05); the values of PI and RI in MCA of CG were significantly higher than those in PIH group (P < 0.05); after NI, PSV, PI, RI, and AI values in the OG were significantly lower than those in the RG (P < 0.05); the health status, trunk pain, social function, emotional function, vitality, mental health, and total score of the OG were significantly higher than that of the RG (P < 0.05), which indicated that ultrasonic detection based on ML algorithm can be used in monitoring the changes of hemodynamic indexes of PIH pregnant women during NI, and the NI based on the Omaha system can effectively improve the QOL of PIH pregnant women.