Assessing Pancreatic Fat and Its Correlation with Liver Fat in Suspected MASLD Patients Using Advanced Deep Learning Techniques from MRI Images

• DOI: 10.3390/app142411924
• Publication Date: 2024-12-20T09:07:40Z
• Authors: Hay Ching Cherrie...; Juan Pablo Meneses; Nirusha Surendran; Numan Kutaiba; Yasmeen George; Enes Makalic; Sergio Uribe
• Keywords: Technology; proton density fat fraction; QH301-705.5; T; Physics; QC1-999; Engineering (General). Civil engineering (General); pancreatic steatosis; deep learning models; chemical-shift-encoded MRI; Chemistry; 03 medical and health sciences; metabolic-dysfunction-associated steatotic fatty disease; 0302 clinical medicine; TA1-2040; Biology (General); QD1-999

Pancreatic steatosis and metabolic-dysfunction-associated steatotic liver disease are characterised by fat accumulation in abdominal organs, but their correlation remains inconclusive. Recently proposed deep learning (DL) for proton density fraction (PDFF) estimation, which quantifies organ fat, has primarily been assessed quantifying fat. This study aims to validate DL models pancreatic PDFF quantification compare pancreas content. We evaluated three models—Non-Linear Variables Neural Network (NLV-Net), U-Net, Multi-Decoder Water-Fat separation Network—against a reference measured using graph-cut-based method. NLV-Net showed strong (Spearman rho) with the six-echo head (slope: 1.02, rho: 0.95) body 1.04, 0.94) moderate three-echo 0.44, 0.40) 0.49, 0.34). Weak correlations were found between graph cut −0.041, −0.12) images 0.0014, 0.073) −0.053, −0.014, −0.033). In conclusion, best agreement quantification, no was