Options
Towards automated age estimation of young individuals : a new computer-based approach using 3D knee MRI
Citation Link: https://doi.org/10.15480/882.2813
Publikationstyp
Doctoral Thesis
Date Issued
2020-06
Sprache
English
Author(s)
Advisor
Referee
Title Granting Institution
Technische Universität Hamburg
Place of Title Granting Institution
Hamburg
Examination Date
2020-02-28
Institut
TORE-DOI
TORE-URI
Citation
Zuerst veröff.: ISBN 978-3-8440-7400-0 (2020)
Publisher DOI
Publisher
Shaker Verlag
The purpose of this work is to investigate the age estimation of living individuals on the basis of MRI sequences of the knee. Using a large data collective of young male subjects, a new AI-based approach is developed to automatically detect bone structures in the image and to learn the chronological age based on this information. With a mean absolute error (MAE) of 0.69 ± 0.47 years for the prediction of the age and an accuracy of 90.9% for the majority classification on the 18-year-limit, the results demonstrate the potential of this approach.
Subjects
automated age estimation
magnetic resonance imaging (MRI)
knee
machine learning
convolutional neural network (CNN)
segmentation
DDC Class
570: Biowissenschaften, Biologie
Loading...
Name
3-8440-7400-7_DOK Auf der Mauer_v2.pdf
Size
25.89 MB
Format
Adobe PDF