20 January 2025 in IEEE Access
To address the issue of inaccurate segmentation caused by blurred edges and strong noise in thyroid nodule ultrasound images, an image segmentation method based on an improved Swin Transformer is proposed. First, depthwise convolutional layers are integrated into the encoder/decoder of the Swin Transformer to enhance global-local feature representations. Second, a multi-scale feature fusion module is introduced through skip connections between the encoder and decoder to improve information flow and feature integration. Additionally, a multi-level patch embedding convolution is designed to enable layer-by-layer feature extraction from coarse to fine levels. Experimental results show that the proposed method achieves superior segmentation accuracy compared to state-of-the-art methods such as Attention U-Net, with Dice scores of 82.26% and 78.64% and IoU values of 73.00% and 67.93% on the TN3K and DDTI datasets, respectively.
25 April 2024 in Engineering Applications of Artificial Intelligence
2 April 2025 in Journal of Imaging Informatics in Medicine
21 April 2023 in Biomedical Optics Express
Because pigmented skin lesion image classification based on manually designed convolutional neural networks (CNNs) requires abundant experience in neural network design and considerable parameter tuning, we proposed the macro operation mutation-based neural architecture search (OM-NAS) approach in order to automatically build a CNN for image classification of pigmented skin lesions. We first used an improved search space that was oriented toward cells and contained micro and macro operations. The macro operations include InceptionV1, Fire and other well-designed neural network modules. During the search process, an evolutionary algorithm based on macro operation mutation was employed to iteratively change the operation type and connection mode of parent cells so that the macro operation was inserted into the child cell similar to the injection of virus into host DNA. Ultimately, the searched best cells were stacked to build a CNN for the image classification of pigmented skin lesions, which was then assessed on the HAM10000 and ISIC2017 datasets. The test results showed that the CNN built with this approach was more accurate than or almost as accurate as state-of-the-art (SOTA) approaches such as AmoebaNet, InceptionV3 + Attention and ARL-CNN in terms of image classification. The average sensitivity of this method on the HAM10000 and ISIC2017 datasets was 72.4% and 58.5%, respectively.
24 January 2018 in Signal, Image and Video Processing
28 February 2013 in International Journal of Digital Content Technology and its Applications