Refine
Document Type
- Article (4)
Language
- English (4)
Has Fulltext
- yes (4)
Is part of the Bibliography
- no (4)
Keywords
- CNN (2)
- Alzheimer’s disease (1)
- Computational biology and bioinformatics (1)
- Diseases (1)
- GCN (1)
- Image processing (1)
- Infectious diseases (1)
- abnormality detection (1)
- attention module (1)
- computer-aided diagnosis (1)
Institute
This study proposes a novel multi-network architecture consisting of a multi-scale convolution neural network (MSCNN) with fully connected graph convolution network (GCN), named MSCNN-GCN, for the detection of musculoskeletal abnormalities via musculoskeletal radiographs. To obtain both detailed and contextual information for a better description of the characteristics of the radiographs, the designed MSCNN contains three subnetwork sequences (three different scales). It maintains high resolution in each sub-network, while fusing features with different resolutions. A GCN structure was employed to demonstrate global structure information of the images. Furthermore, both the outputs of MSCNN and GCN were fused through the concat of the two feature vectors from them, thus making the novel framework more discriminative. The effectiveness of this model was verified by comparing the performance of radiologists and three popular CNN models (DenseNet169, CapsNet, and MSCNN) with three evaluation metrics (Accuracy, F1 score, and Kappa score) using the MURA dataset (a large dataset of bone X-rays). Experimental results showed that the proposed framework not only reached the highest accuracy, but also demonstrated top scores on both F1 metric and kappa metric. This indicates that the proposed model achieves high accuracy and strong robustness in musculoskeletal radiographs, which presents strong potential for a feasible scheme with intelligent medical cases.
Alzheimer’s disease (AD) is the most prevalent neurodegenerative disease causing dementia and poses significant health risks to middle-aged and elderly people. Brain magnetic resonance imaging (MRI) is the most widely used diagnostic method for AD. However, it is challenging to collect sufficient brain imaging data with high-quality annotations. Weakly supervised learning (WSL) is a machine learning technique aimed at learning effective feature representation from limited or low-quality annotations. In this paper, we propose a WSL-based deep learning (DL) framework (ADGNET) consisting of a backbone network with an attention mechanism and a task network for simultaneous image classification and image reconstruction to identify and classify AD using limited annotations. The ADGNET achieves excellent performance based on six evaluation metrics (Kappa, sensitivity, specificity, precision, accuracy, F1-score) on two brain MRI datasets (2D MRI and 3D MRI data) using fine-tuning with only 20% of the labels from both datasets. The ADGNET has an F1-score of 99.61% and sensitivity is 99.69%, outperforming two state-of-the-art models (ResNext WSL and SimCLR). The proposed method represents a potential WSL-based computer-aided diagnosis method for AD in clinical practice.
The intergroup sensitivity effect in mergers and acquisitions: testing the role of merger motives
(2021)
Research has shown that people are more defensive to criticism when it stems from an outgroup member, compared to an ingroup member (the intergroup sensitivity effect: ISE). We conducted two online vignette experiments to examine the ISE in the context of an organizational merger and the role of merger motives for the ISE. We predicted that the ISE would also emerge in mergers and acquisitions (M&As), but people would respond less negatively to criticism from the outgroup when the motive for the merger is described as achieving synergies rather than growth. In Experiment 1 (N = 452), which did not mention any motives behind the acquisition, a significant ISE emerged. Experiment 2 (N = 587) again showed an ISE regardless of the merger motive. In both experiments, the ISE was mediated by perceptions of the outgroup criticism as less legitimate and constructive. Overall, this research points to the intergroup sensitivity effect as a relevant phenomenon during post-merger integration.
Coronavirus disease 2019 (COVID-19) is a global pandemic posing significant health risks. The diagnostic test sensitivity of COVID-19 is limited due to irregularities in specimen handling. We propose a deep learning framework that identifies COVID-19 from medical images as an auxiliary testing method to improve diagnostic sensitivity. We use pseudo-coloring methods and a platform for annotating X-ray and computed tomography images to train the convolutional neural network, which achieves a performance similar to that of experts and provides high scores for multiple statistical indices (F1 scores > 96.72% (0.9307, 0.9890) and specificity >99.33% (0.9792, 1.0000)). Heatmaps are used to visualize the salient features extracted by the neural network. The neural network-based regression provides strong correlations between the lesion areas in the images and five clinical indicators, resulting in high accuracy of the classification framework. The proposed method represents a potential computer-aided diagnosis method for COVID-19 in clinical practice.