Loading [MathJax]/extensions/TeX/upgreek.js
Brain Devlopment Age Prediction Using Convolutnal Neural Network on Pediatrics Brain Ct Mages | IEEE Conference Publication | IEEE Xplore

Brain Devlopment Age Prediction Using Convolutnal Neural Network on Pediatrics Brain Ct Mages


Abstract:

Progress of normal development of the brain, such as underdevelopment or prematurity, is evaluated as one of the indices to diagnose pediatric brain diseases using brain ...Show More

Abstract:

Progress of normal development of the brain, such as underdevelopment or prematurity, is evaluated as one of the indices to diagnose pediatric brain diseases using brain images. However, there are no methods to quantitatively estimate the degree of brain development, and the current diagnosis is based on the experience of physicians. This study proposes a method to predict brain development age on pediatric brain CT images. The method segments the cranial region from CT images and aligns the pose and orientation. We propose a novel network model which extracts features from the CT images with 3D convolutional neural network (CNN) and predicts the brain development age with fully-connected layers. Its performance was evaluated using 60 normally developing newborns between the ages of 0 and 3 years. The root-mean-square-error between predicted age and chronological age was 7.80 in months and the correlation coefficient was 0.801.
Date of Conference: 04-05 December 2021
Date Added to IEEE Xplore: 24 March 2022
ISBN Information:

ISSN Information:

Conference Location: Adelaide, Australia

References

References is not available for this document.