Pseudo-domains in imaging data improve prediction of future disease\n status in multi-center studies

2021/11/15 by Matthias Perkonigg, Peter Mesenbrink, Perkonigg, Matthias +9
Biochemistry, Genetics and Molecular Biology · Mathematics · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Gene expression and cancer classification #Image and Video Processing (eess.IV) #Liver Disease Diagnosis and Treatment #Machine Learning (cs.LG) #Radiomics and Machine Learning in Medical Imaging #Statistical Methods and Inference #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2111.07634

openalex publication_date 2021/11/15 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In multi-center randomized clinical trials imaging data can be diverse due to\nacquisition technology or scanning protocols. Models predicting future outcome\nof patients are impaired by this data heterogeneity. Here, we propose a\nprediction method that can cope with a high number of different scanning sites\nand a low number of samples per site. We cluster sites into pseudo-domains\nbased on visual appearance of scans, and train pseudo-domain specific models.\nResults show that they improve the prediction accuracy for steatosis after 48\nweeks from imaging data acquired at an initial visit and 12-weeks follow-up in\nliver disease\n

Citations

• Least squares quantization in PCM

Related

• Multi-domain improves out-of-distribution and data-limited scenarios for medical image analysis
• Domain Generalizer: A Few-shot Meta Learning Framework for Domain\n Generalization in Medical Imaging
• When Unseen Domain Generalization is Unnecessary? Rethinking Data Augmentation
• Domain Generalization by Learning from Privileged Medical Imaging Information
• A Strong Baseline for Domain Adaptation and Generalization in Medical Imaging