Designing an Improved Deep Learning-based Model for COVID-19 Recognition in Chest X-ray Images: A Knowledge Distillation Approach

Abstract

<title>Abstract</title> Background and Objectives: COVID-19 has adversely affected humans and societies in different aspects. Numerous people have perished due to inaccurate COVID-19 identification and, consequently, a lack of appropriate medical treatment. Numerous solutions based on manual and automatic feature extraction techniques have been investigated to address this issue by researchers worldwide. Typically, automatic feature extraction methods, particularly deep learning models, necessitate a powerful hardware system to perform the necessary computations. Unfortunately, many institutions and societies cannot benefit from these advancements due to the prohibitively high cost of high-quality hardware equipment. As a result, this study focused on two primary goals: first, lowering the computational costs associated with running the proposed model on embedded devices, mobile devices, and conventional computers; and second, improving the model's performance in comparison to previously published methods (at least performs on par with state of the art models) in order to ensure its performance and accuracy for the medical recognition task. Methods This study used two neural networks to improve feature extraction from our dataset: VGG19 and ResNet50V2. Both of these networks are capable of providing semantic features from the nominated dataset. Streaming is a fully connected classifier layer that feeds richer features, therefore feature vectors of these networks have been merged, and this action resulted in satisfactory classification results for normal and COVID-19 cases. On the other hand, these two networks have many layers and require a significant amount of computation. To this end, An alternative network was considered, namely MobileNetV2, which excels at extracting semantic features while requiring minimal computation on mobile and embedded devices. Knowledge distillation (KD) was used to transfer knowledge from the teacher network (concatenated ResNet50V2 and VGG19) to the student network (MobileNetV2) to improve MobileNetV2 performance and to achieve a robust and accurate model for the COVID-19 identification task from chest X-ray images. Results Pre-trained networks were used to provide a more useful starting point for the COVID-19 detection task. Additionally, a 5-fold cross-validation technique was used on both the teacher and student networks to evaluate the proposed method's performance. Finally, the proposed model achieved 98.8% accuracy in detecting infectious and normal cases. Conclusion The study results demonstrate the proposed method's superior performance. With the student model achieving acceptable accuracy and F1-score using cross-validation technique, it can be concluded that this network is well-suited for conventional computers, embedded systems, and clinical experts' cell phones.

Ähnliche Arbeiten