The Intrinsic Manifolds of Radiological Images and their Role in Deep Learning

Abstract

The manifold hypothesis is a core mechanism behind the success of deep learning, so understanding the intrinsic manifold structure of image data is central to studying how neural networks learn from the data. Intrinsic dataset manifolds and their relationship to learning difficulty have recently begun to be studied for the common domain of natural images, but little such research has been attempted for radiological images. We address this here. First, we compare the intrinsic manifold dimensionality of radiological and natural images. We also investigate the relationship between intrinsic dimensionality and generalization ability over a wide range of datasets. Our analysis shows that natural image datasets generally have a higher number of intrinsic dimensions than radiological images. However, the relationship between generalization ability and intrinsic dimensionality is much stronger for medical images, which could be explained as radiological images having intrinsic features that are more difficult to learn. These results give a more principled underpinning for the intuition that radiological images can be more challenging to apply deep learning to than natural image datasets common to machine learning research. We believe rather than directly applying models developed for natural images to the radiological imaging domain, more care should be taken to developing architectures and algorithms that are more tailored to the specific characteristics of this domain. The research shown in our paper, demonstrating these characteristics and the differences from natural images, is an important first step in this direction.

arXiv commentpreprint version, accepted for MICCAI 2022 (25th International Conference on Medical Image Computing and Computer Assisted Intervention). 8 pages (+ author names + references + supplementary), 4 figures. Code available at https://github.com/mazurowski-lab/radiologyintrinsicmanifolds

References in corpus (7)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Estimating the intrinsic dimension of datasets by a minimal neighborhood information
• A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets
• MURA: Large Dataset for Abnormality Detection in Musculoskeletal Radiographs
• Intrinsic dimension of data representations in deep neural networks
• Scikit-dimension: a Python package for intrinsic dimension estimation

Cited by in corpus (3)

• Evaluating Visual Explanations of Attention Maps for Transformer-based Medical Imaging
• Exploring Patient Data Requirements in Training Effective AI Models for MRI-based Breast Cancer Classification
• Density Estimation via Binless Multidimensional Integration