Deep learning-based muscle segmentation for reproducible fat and water quantification using magnetic resonance imaging – applications to normal ageing and muscular dystrophy

Professor Bae, Kyongtae Tyler   (Principal Investigator (PI))

Dou Qi   (Co-Investigator)

Dr Chang Hing Chiu   (Co-Investigator)

Professor Cao Peng   (Co-Investigator)

Ho Siu Lun Ronnie   (Co-Investigator)

Lee Chi Nam   (Co-Investigator)

Professor Javed Asif   (Co-Investigator)

36

2022-10-01

Deep learning-based muscle segmentation for reproducible fat and water quantification using magnetic resonance imaging – applications to normal ageing and muscular dystrophy

artificial intelligence, deep learning, muscular dystrophy, ageing, muscle quantification

Medicine, Dentistry and Health

09202366

Health and Medical Research Fund - Full Grant

2022

On-going

1) To develop DL segmentation for automated individual thigh muscle segmentation using MRI. 2) To establish normal fat-water quantification values in healthy ageing population.3) To prospectively assess the accuracy of DL based volumetric quantification compared with traditional assessment with limited axial MRI slices and/or visual analysis.4) Using aged-specific normative values established in objective 3, we will prospectively evaluate DL method in patients with myopathies with traditional clinical assessment as comparison.