Quantitative Imaging Principles Improves Medical Image Learning
Preprint
Abstract
Fundamental differences between natural and medical images have recently favored the use of self-supervised learning (SSL) over ImageNet transfer learning for medical image applications. Differences between image types are primarily due to the imaging modality and medical images utilize a wide range of physics based techniques while natural images are captured using only visible light. While many have demonstrated that SSL on medical images has resulted in better downstream task performance, our work suggests that more performance can be gained. The scientific principles which are used to acquire medical images are not often considered when constructing learning problems. For this reason, we propose incorporating quantitative imaging principles during generative SSL to improve image quality and quantitative biological accuracy. We show that this training schema results in better starting states for downstream supervised training on limited data. Our model also generates images that validate on clinical quantitative analysis software.Keywords
All Related Versions
This publication has 22 references indexed in Scilit:
- Pixel-wise body composition prediction with a multi-task conditional generative adversarial networkJournal of Biomedical Informatics, 2021
- A scoping review of transfer learning research on medical image analysis using ImageNetComputers in Biology and Medicine, 2021
- Prediction of trabecular bone architectural features by deep learning models using simulated DXA imagesBone Reports, 2020
- Deep learning of lumbar spine X-ray for osteopenia and osteoporosis screening: A multicenter retrospective cohort studyBone, 2020
- MedGAN: Medical image translation using GANsComputerized Medical Imaging and Graphics, 2020
- Simulating Dual-Energy X-Ray Absorptiometry in CT Using Deep-Learning Segmentation CascadeJournal of the American College of Radiology, 2019
- Not-so-supervised: A survey of semi-supervised, multi-instance, and transfer learning in medical image analysisMedical Image Analysis, 2019
- Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equationsJournal of Computational Physics, 2019
- Body composition by DXABone, 2017
- Measurement of lean body mass and total body fat using dual photon absorptiometryMetabolism, 1986