Early Prediction of Hydrocephalus Using Data Mining Techniques

Abstract

Hydrocephalus is a condition which is characterized by head enlargement in infants due to enlargement of brain ventricles. An excess of fluid secretion and collection of fluid within the brain cavities are treated as Hydrocephalus. The extra fluid puts stress on the brain and can damage the brain. Hence, the increase in the fluid level in the brain's cavities may increase intracranial pressure and lead to brain damage. It is most usual in infants on children and rarely in the adult age group. Children often have a full life span if hydrocephalus is early detected and treated. This paper presents various data mining techniques and used to find out the disease in an early manner. Magnetic resonance imaging is one of the detection tools which is used to predict the disease properly. This approach includes the basic four data mining processes, namely preprocessing, segmentation, feature extraction, and classification as stage-by-stage manner using MRI dataset. Along with this process, the tree augmented Naïve Bayes nearest neighbor (TANNN) algorithm is also implemented to improve the accuracy in detecting the disease and also gave the best detection rate. The TANNN algorithm may provide the best results in diplomatic, uniqueness, perfection, and overall running time. The first stage in the data mining technique is preprocessing, which converts the original data into a useful format. The second stage is a key technique, and it groups the original data into possible divisions according to its category. The third stage is feature extraction, which is used to extract the needed data from the source. The fourth stage is the classification that appoints data in a collection to destination categories or data groups. This paper also concentrates on image mining, which includes experiments in image elements such as texture, shape, and size. Image classification is an important task in the field of medicine and technology. This helps the radiologist in the process of diagnosing hydrocephalus.