Estimation of brain age from MRI images using K-nearest neighbour and compared with recurrent neural network to improve the accuracy

The intention of this study is to improve accuracy in the brain age estimation system from MRI images by using K-Nearest Neighbour (KNN) compared with the Recurrent Neural Network (RNN). Improved the accuracy in brain age estimation system from MRI images by comparing KNN with RNN. This analysis is done in the SPSS software for better outcome. There were two groups, group 1 and group 2, and each group had 20 samples. The G power and alpha values obtained as 80% and 0.001. The accuracy performance is increased by the KNN with RNN for brain age estimation of 92.58% and 89.70% respectively. In this, the independent sample t-test value is obtained as 0.001 (p < 0.05). It demonstrates that the two methods have statistically different results. This concludes that KNN has better accuracy performance by comparison to the RNN in brain age estimation. The KNN has high accuracy to estimate the brain age.