Augmentation strategies for an imbalanced learning problem on a novel COVID-19 severity dataset

Abstract

Since the beginning of the COVID-19 pandemic, many different machine learning models have been developed to detect and verify COVID-19 pneumonia based on chest X-ray images. Although promising, binary models have only limited implications for medical treatment, whereas the prediction of disease severity suggests more suitable and specific treatment options. In this study, we publish severity scores for the 2358 COVID-19 positive images in the COVIDx8B dataset, creating one of the largest collections of publicly available COVID-19 severity data. Furthermore, we train and evaluate deep learning models on the newly created dataset to provide a first benchmark for the severity classification task. One of the main challenges of this dataset is the skewed class distribution, resulting in undesirable model performance for the most severe cases. We therefore propose and examine different augmentation strategies, specifically targeting majority and minority classes. Our augmentation strategies show significant improvements in precision and recall values for the rare and most severe cases. While the models might not yet fulfill medical requirements, they serve as an appropriate starting point for further research with the proposed dataset to optimize clinical resource allocation and treatment.

A Machine Learning Model for the Prediction of COVID-19 Severity Using RNA-Seq, Clinical, and Co-Morbidity Data

“…We used Adaptive Synthetic Sampling (ADASYN) to oversample the minority class and address the class imbalance problem [16]. ADASYN mitigates this issue by adaptively generating synthetic samples for the minority class based on the local density distribution of existing instances [17]. The algorithm works mainly in four steps: (1) the data distribution analysis of all the classes, (2) the density estimation and identification of k-nearest neighbors of all instances in the minority classes, (3) the difficulty level measurement of minority and majority class instances, and (4) adaptive sampling based on the difficulty ratio to determine the number of synthetic samples needed for each minority class instance.…”

A Critical Assessment of Generative Models for Synthetic Data Augmentation on Limited Pneumonia X-ray Data

“…We employ an augmentation pipeline for all classification models to increase image variations and reduce overfitting during model training, which is common for many image domains [53][54][55]. This pipeline was inspired by the winning solution to the 2021 SIIM-FISABIO-RSNA Machine Learning COVID-19 Challenge [56] and is shown in Table 4.…”