Risk Assessment of hematoma expansion in patients with hemorrhagic Stroke based on Stochastic Forest Prediction Model

Yuwei Xiong; Xiaoyi Feng; Feiran Gao

doi:10.62051/1r3hce57

Authors

Yuwei Xiong
Xiaoyi Feng
Feiran Gao

DOI:

https://doi.org/10.62051/1r3hce57

Keywords:

Hemorrhagic stroke; Hematoma expansion; Predictive model; Support vector machine (SVM); Random Forest.

Abstract

This paper discusses the prediction model of hematoma expansion probability in patients with hemorrhagic stroke. Hemorrhagic stroke is an acute and critical neurological disease. High mortality and disability make it a major challenge in the field of public health. Hematoma dilatation and peri-hematoma edema are important factors leading to secondary brain injury. therefore, accurate prediction of the risk of hematoma expansion in patients with hemorrhagic stroke has important guiding significance for clinical diagnosis and treatment. First of all, this study carries on the matching integration and abnormal processing of the patient data in order to improve the data quality. Then, a classification model based on support vector machine (SVM) and random forest is constructed to predict the probability of hematoma expansion. Through the comparative analysis of the classification accuracy, it is found that the prediction effect of the classification model based on random forest is better than that based on SVM. This result provides clinicians with a more accurate prediction tool, which is helpful to identify high-risk patients early, reduce the incidence of secondary brain injury, and improve the quality of life and prognosis of patients. This study provides a new idea for the clinical diagnosis, treatment and research direction of hemorrhagic stroke, and has important theoretical and practical value.

Downloads

Download data is not yet available.

References

LI Jianbo, GENG Wei, ZHOU Tao, et al. Predictive modeling of hematoma expansion risk in patients with cerebral hemorrhage based on computed tomography angiographic features [J]. Journal of practical clinical medicine, 2022 (026 - 008).

B H Z A, B Z, B Z S, et al. Machine learning-based modified BAT score in predicting hematoma enlargement after spontaneous intracerebral hemorrhage [J]. 2021.

XU Lei, GE Huaizhi, ZHANG Zhijing, et al. The value of predicting hematoma enlargement after spontaneous intracerebral hemorrhage based on clinical features and modeling of imaging histology features in plain CT [J]. Journal of Wenzhou Medical University, 2021, 51 (10): 6.

FAN Jingqian, JIANG Yuyan. Research progress on the mechanism of edema formation around hematoma after cerebral hemorrhage [J]. Shandong Medicine, 2021, 61 (2): 3.

Cortes C, Vapnik V. Support-vector networks [J]. Machine learning, 1995, 20 (3): 273 - 297.

Cui CJ. Research on optical image target recognition technology based on SVM and CNN [D]. Harbin Institute of Technology, 2022.

Ahmed M Y O Z M. Research on privacy-preserving SVM model for distributed medical data analysis [D]. University of Electronic Science and Technology, 2018.

Yin J P. Research on support vector machine kernel function and key parameters selection [D]. Harbin Institute of Technology, 2016.

Li Lin. Disease prediction analysis based on optimized random forest mixture model [D]. North China University of Water Resources and Hydropower, 2022.

Sounak C, Mohammed K, Mihail P. Predicting disease risks from highly imbalanced data using random forest [J]. BMC Medical Informatics and Decision Making, 2011, 11 (1).