RESUMO
In the context of the global coronavirus pandemic, different deep learning solutions for infected subject detection using chest X-ray images have been proposed. However, deep learning models usually need large labelled datasets to be effective. Semi-supervised deep learning is an attractive alternative, where unlabelled data is leveraged to improve the overall model's accuracy. However, in real-world usage settings, an unlabelled dataset might present a different distribution than the labelled dataset (i.e. the labelled dataset was sampled from a target clinic and the unlabelled dataset from a source clinic). This results in a distribution mismatch between the unlabelled and labelled datasets. In this work, we assess the impact of the distribution mismatch between the labelled and the unlabelled datasets, for a semi-supervised model trained with chest X-ray images, for COVID-19 detection. Under strong distribution mismatch conditions, we found an accuracy hit of almost 30%, suggesting that the unlabelled dataset distribution has a strong influence in the behaviour of the model. Therefore, we propose a straightforward approach to diminish the impact of such distribution mismatch. Our proposed method uses a density approximation of the feature space. It is built upon the target dataset to filter out the observations in the source unlabelled dataset that might harm the accuracy of the semi-supervised model. It assumes that a small labelled source dataset is available together with a larger source unlabelled dataset. Our proposed method does not require any model training, it is simple and computationally cheap. We compare our proposed method against two popular state of the art out-of-distribution data detectors, which are also cheap and simple to implement. In our tests, our method yielded accuracy gains of up to 32%, when compared to the previous state of the art methods. The good results yielded by our method leads us to argue in favour for a more data-centric approach to improve model's accuracy. Furthermore, the developed method can be used to measure data effectiveness for semi-supervised deep learning model training.
RESUMO
The implementation of deep learning-based computer-aided diagnosis systems for the classification of mammogram images can help in improving the accuracy, reliability, and cost of diagnosing patients. However, training a deep learning model requires a considerable amount of labelled images, which can be expensive to obtain as time and effort from clinical practitioners are required. To address this, a number of publicly available datasets have been built with data from different hospitals and clinics, which can be used to pre-train the model. However, using models trained on these datasets for later transfer learning and model fine-tuning with images sampled from a different hospital or clinic might result in lower performance. This is due to the distribution mismatch of the datasets, which include different patient populations and image acquisition protocols. In this work, a real-world scenario is evaluated where a novel target dataset sampled from a private Costa Rican clinic is used, with few labels and heavily imbalanced data. The use of two popular and publicly available datasets (INbreast and CBIS-DDSM) as source data, to train and test the models on the novel target dataset, is evaluated. A common approach to further improve the model's performance under such small labelled target dataset setting is data augmentation. However, often cheaper unlabelled data is available from the target clinic. Therefore, semi-supervised deep learning, which leverages both labelled and unlabelled data, can be used in such conditions. In this work, we evaluate the semi-supervised deep learning approach known as MixMatch, to take advantage of unlabelled data from the target dataset, for whole mammogram image classification. We compare the usage of semi-supervised learning on its own, and combined with transfer learning (from a source mammogram dataset) with data augmentation, as also against regular supervised learning with transfer learning and data augmentation from source datasets. It is shown that the use of a semi-supervised deep learning combined with transfer learning and data augmentation can provide a meaningful advantage when using scarce labelled observations. Also, we found a strong influence of the source dataset, which suggests a more data-centric approach needed to tackle the challenge of scarcely labelled data. We used several different metrics to assess the performance gain of using semi-supervised learning, when dealing with very imbalanced test datasets (such as the G-mean and the F2-score), as mammogram datasets are often very imbalanced. Graphical Abstract Description of the test-bed implemented in this work. Two different source data distributions were used to fine-tune the different models tested in this work. The target dataset is the in-house CR-Chavarria-2020 dataset.
Assuntos
Diagnóstico por Computador , Aprendizado de Máquina Supervisionado , Costa Rica , Diagnóstico por Computador/métodos , Humanos , Mamografia , Reprodutibilidade dos TestesRESUMO
Developers proposing new machine learning for health (ML4H) tools often pledge to match or even surpass the performance of existing tools, yet the reality is usually more complicated. Reliable deployment of ML4H to the real world is challenging as examples from diabetic retinopathy or Covid-19 screening show. We envision an integrated framework of algorithm auditing and quality control that provides a path towards the effective and reliable application of ML systems in healthcare. In this editorial, we give a summary of ongoing work towards that vision and announce a call for participation to the special issue Machine Learning for Health: Algorithm Auditing & Quality Control in this journal to advance the practice of ML4H auditing.
Assuntos
Algoritmos , Aprendizado de Máquina , Controle de Qualidade , HumanosRESUMO
In this work we implement a COVID-19 infection detection system based on chest X-ray images with uncertainty estimation. Uncertainty estimation is vital for safe usage of computer aided diagnosis tools in medical applications. Model estimations with high uncertainty should be carefully analyzed by a trained radiologist. We aim to improve uncertainty estimations using unlabelled data through the MixMatch semi-supervised framework. We test popular uncertainty estimation approaches, comprising Softmax scores, Monte-Carlo dropout and deterministic uncertainty quantification. To compare the reliability of the uncertainty estimates, we propose the usage of the Jensen-Shannon distance between the uncertainty distributions of correct and incorrect estimations. This metric is statistically relevant, unlike most previously used metrics, which often ignore the distribution of the uncertainty estimations. Our test results show a significant improvement in uncertainty estimates when using unlabelled data. The best results are obtained with the use of the Monte Carlo dropout method.
RESUMO
A key factor in the fight against viral diseases such as the coronavirus (COVID-19) is the identification of virus carriers as early and quickly as possible, in a cheap and efficient manner. The application of deep learning for image classification of chest X-ray images of COVID-19 patients could become a useful pre-diagnostic detection methodology. However, deep learning architectures require large labelled datasets. This is often a limitation when the subject of research is relatively new as in the case of the virus outbreak, where dealing with small labelled datasets is a challenge. Moreover, in such context, the datasets are also highly imbalanced, with few observations from positive cases of the new disease. In this work we evaluate the performance of the semi-supervised deep learning architecture known as MixMatch with a very limited number of labelled observations and highly imbalanced labelled datasets. We demonstrate the critical impact of data imbalance to the model's accuracy. Therefore, we propose a simple approach for correcting data imbalance, by re-weighting each observation in the loss function, giving a higher weight to the observations corresponding to the under-represented class. For unlabelled observations, we use the pseudo and augmented labels calculated by MixMatch to choose the appropriate weight. The proposed method improved classification accuracy by up to 18%, with respect to the non balanced MixMatch algorithm. We tested our proposed approach with several available datasets using 10, 15 and 20 labelled observations, for binary classification (COVID-19 positive and normal cases). For multi-class classification (COVID-19 positive, pneumonia and normal cases), we tested 30, 50, 70 and 90 labelled observations. Additionally, a new dataset is included among the tested datasets, composed of chest X-ray images of Costa Rican adult patients.