Xiaolan Li ^* Fen Kang Xiaojing Li

• Anhui Medical University, Hefei, China

The study proposes an intelligent sampling and feature selection method based on F-Scores optimization and designs an integrated prediction model, incorporating genetic algorithms and multiple heterogeneous learners in order to fully utilize the potential value of medical data, prevent misdiagnosis due to inexperience of related personnel and so on, and improve the quality of treatment. The study's improved intelligent feature selection method yielded several features of pregnancy-related hypertension, including phosphor dehydrogenase deficiency, body mass index, gestational urinary proteins, vascular endothelial growth factor receptor 1, placental growth factor, thalassemia, and a family history of diabetes mellitus or hypertension. The study's model had the best recall, F-Score, and area under the curve when compared to other mainstream models; these values were 0.768, 0.728, and 0.832, respectively. The area under the curve for both early and late haircuts in clinical application was maximum at 0.996 and 0.792, respectively, when measured by the ratio of vascular endothelial growth factor receptor 1 compared to placental growth factor. The above results indicated that the intelligent gestational hypertension diagnosis and prediction method proposed in the study had excellent performance and could be better applied in the teaching and clinical practice of obstetrics and gynecology, which is helpful for promoting intelligent medical diagnosis in China.