Adopting Random Forest For Predicting The Risk Of Cerebrovascular Disease Using The BraVa Datasets

Characterization of complex architecture of cerebral arteries across a representative sample of human population is important for diagnosing,analyzing and prediction of pathological states.Brain arterial vasculature can be visualized by magnetic resonance angiography.The advent of automatic tracing and reconstruction technology has led to a surge in the number of neurons 3D reconstruction data and consequently the Neuromorphology research.However,the lack of machine-driven annotation schema to automatically detect and make prediction of a possible disorder or risk of a subject being affected with a cerebrovascular disease,using the neuromorphological measurement values is still a hinderance in this branch of science.Random Forest(RF)is a commonly used method of machine learning that demonstrates competitive success in various fields,including biological science,economics,chemical engineering,Agro Sciences,medical research,etc.In critical application such as diagnosing and prediction of pathological states.It is a technique which comprise of a simple tree predictor,such that when a set of predictor values is given as input,each tree produces a response.This thesis provides a comprehensive research in the field of machine learning where a decision tree based random algorithm,random forest is used to make prediction of the possibility of the existence of a cerebrovascular disease in 44 subjects curated from the BraVa datasets.Towards analyzing and prediction of cerebrovascular diseases,this thesis makes use of the statistical tool(SPSS)for realizing the independence and correlation between the various metrics that could contribute to a subject been diagnosed with a cerebrovascular disease.A summary statistic for the various scaler parameters that characterize the entire vascular structure with regards to overall size,branch features and bifurcation angles and symmetry are computed.Overall size variability of the data used in this study was similar to the values reported for other parameters of human body size.For instance,the coefficients of variation for total number of branches and total length were approximately between 0.13 and 0.25,and associated ranges between 68%-157% of the means of the respective parameters.An analysis of the various metrics of the overall vascular size revealed a significant correlation especially between Total Length and Total Number of Branches(R = 0.829,p = 0.000).A statistically significant difference also existed between Total Length and MBO(R = 0.323,p<0.035),Height(R = 0.354,p<0.02),Depth(R = 0.438,p<0.003)and Tortuosity(R = 1.000,p=0.000).A statistically significant difference was also observed between MBO and PA(R = 0.323,p<0.035),BPL(R =-0.308,p<0.045)and Tortuosity(R = 0.323,p<0.032).Studies have shown that these parameters are critical in the determination of a person’s risk of getting cerebrovascular diseases.The random forest algorithm was then used as a second tool to predict the risk of a subject being affected be cerebrovascular disease.Metrics like Age,Contraction,Tortuosity,mean bifurcation Angle,mean bifurcation tilt which has implication of a cerebrovascular disease diagnosis according to study was used as the input for the random forest algorithm.The BraVa dataset which is the main datasets for this work was used to train the algorithm and a prediction of either “risky” or “Not risky” with a high accuracy of 100% was recorded.To further test the algorithm,a second datasets from the from the diabetes database which has a high number of subjects was also used to test the algorithm and a high accuracy of 90.256% was recorded.In was determined from the results that machine learning based Random Forest algorithm can be adopted as a prediction method especially on bigger dataset of neuromorphological measurements of neurons and it will aid or facilitate accurate prediction of any form of cerebrovascular disease and also aid in accurate medical diagnosis.