Dataset Information

Self-supervised deep learning model for COVID-19 lung CT image segmentation highlighting putative causal relationship among age, underlying disease and COVID-19.

ABSTRACT:

Background

Coronavirus disease 2019 (COVID-19) is very contagious. Cases appear faster than the available Polymerase Chain Reaction test kits in many countries. Recently, lung computerized tomography (CT) has been used as an auxiliary COVID-19 testing approach. Automatic analysis of the lung CT images is needed to increase the diagnostic efficiency and release the human participant. Deep learning is successful in automatically solving computer vision problems. Thus, it can be introduced to the automatic and rapid COVID-19 CT diagnosis. Many advanced deep learning-based computer vison techniques were developed to increase the model performance but have not been introduced to medical image analysis.

Methods

In this study, we propose a self-supervised two-stage deep learning model to segment COVID-19 lesions (ground-glass opacity and consolidation) from chest CT images to support rapid COVID-19 diagnosis. The proposed deep learning model integrates several advanced computer vision techniques such as generative adversarial image inpainting, focal loss, and lookahead optimizer. Two real-life datasets were used to evaluate the model's performance compared to the previous related works. To explore the clinical and biological mechanism of the predicted lesion segments, we extract some engineered features from the predicted lung lesions. We evaluate their mediation effects on the relationship of age with COVID-19 severity, as well as the relationship of underlying diseases with COVID-19 severity using statistic mediation analysis.

Results

The best overall F1 score is observed in the proposed self-supervised two-stage segmentation model (0.63) compared to the two related baseline models (0.55, 0.49). We also identified several CT image phenotypes that mediate the potential causal relationship between underlying diseases with COVID-19 severity as well as the potential causal relationship between age with COVID-19 severity.

Conclusions

This work contributes a promising COVID-19 lung CT image segmentation model and provides predicted lesion segments with potential clinical interpretability. The model could automatically segment the COVID-19 lesions from the raw CT images with higher accuracy than related works. The features of these lesions are associated with COVID-19 severity through mediating the known causal of the COVID-19 severity (age and underlying diseases).

SUBMITTER: Fung DLX

PROVIDER: S-EPMC8312213 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Background and objectiveCoronavirus disease 2019 (COVID-19) is a highly contagious virus spreading all around the world. Deep learning has been adopted as an effective technique to aid COVID-19 detection and segmentation from computed tomography (CT) images. The major challenge lies in the inadequate public COVID-19 datasets. Recently, transfer learning has become a widely used technique that leverages the knowledge gained while solving one problem and applying it to a different but related problem. However, it remains unclear whether various non-COVID19 lung lesions could contribute to segmenting COVID-19 infection areas and how to better conduct this transfer procedure. This paper provides a way to understand the transferability of non-COVID19 lung lesions and a better strategy to train a robust deep learning model for COVID-19 infection segmentation.MethodsBased on a publicly available COVID-19 CT dataset and three public non-COVID19 datasets, we evaluate four transfer learning methods using 3D U-Net as a standard encoder-decoder method. i) We introduce the multi-task learning method to get a multi-lesion pre-trained model for COVID-19 infection. ii) We propose and compare four transfer learning strategies with various performance gains and training time costs. Our proposed Hybrid-encoder Learning strategy introduces a Dedicated-encoder and an Adapted-encoder to extract COVID-19 infection features and general lung lesion features, respectively. An attention-based Selective Fusion unit is designed for dynamic feature selection and aggregation.ResultsExperiments show that trained with limited data, proposed Hybrid-encoder strategy based on multi-lesion pre-trained model achieves a mean DSC, NSD, Sensitivity, F1-score, Accuracy and MCC of 0.704, 0.735, 0.682, 0.707, 0.994 and 0.716, respectively, with better genetalization and lower over-fitting risks for segmenting COVID-19 infection.ConclusionsThe results reveal the benefits of transferring knowledge from non-COVID19 lung lesions, and learning from multiple lung lesion datasets can extract more general features, leading to accurate and robust pre-trained models. We further show the capability of the encoder to learn feature representations of lung lesions, which improves segmentation accuracy and facilitates training convergence. In addition, our proposed Hybrid-encoder learning method incorporates transferred lung lesion features from non-COVID19 datasets effectively and achieves significant improvement. These findings promote new insights into transfer learning for COVID-19 CT image segmentation, which can also be further generalized to other medical tasks.

Project description:Appropriate wound management shortens the healing times and reduces the management costs, benefiting the patient in physical terms and potentially reducing the healthcare system’s economic burden. Among the instrumental measurement methods, the image analysis of a wound area is becoming one of the cornerstones of chronic ulcer management. Our study aim is to develop a solid AI method based on a convolutional neural network to segment the wounds efficiently to make the work of the physician more efficient, and subsequently, to lay the foundations for the further development of more in-depth analyses of ulcer characteristics. In this work, we introduce a fully automated model for identifying and segmenting wound areas which can completely automatize the clinical wound severity assessment starting from images acquired from smartphones. This method is based on an active semi-supervised learning training of a convolutional neural network model. In our work, we tested the robustness of our method against a wide range of natural images acquired in different light conditions and image expositions. We collected the images using an ad hoc developed app and saved them in a database which we then used for AI training. We then tested different CNN architectures to develop a balanced model, which we finally validated with a public dataset. We used a dataset of images acquired during clinical practice and built an annotated wound image dataset consisting of 1564 ulcer images from 474 patients. Only a small part of this large amount of data was manually annotated by experts (ground truth). A multi-step, active, semi-supervised training procedure was applied to improve the segmentation performances of the model. The developed training strategy mimics a continuous learning approach and provides a viable alternative for further medical applications. We tested the efficiency of our model against other public datasets, proving its robustness. The efficiency of the transfer learning showed that after less than 50 epochs, the model achieved a stable DSC that was greater than 0.95. The proposed active semi-supervised learning strategy could allow us to obtain an efficient segmentation method, thereby facilitating the work of the clinician by reducing their working times to achieve the measurements. Finally, the robustness of our pipeline confirms its possible usage in clinical practice as a reliable decision support system for clinicians.