Scholarlink

Carotid Artery Stenosis (CAS) is the condition, where the carotid artery blocks the blood supply to the brain. This can lead to strokes and heart attacks. In general, the CAS caused by the atherosclerosis, where in the arteries build the plaque and block the blood flow. Medical imaging techniques are used to detect the CAS; however, it often faces limitations like high costs and challenges in exposure to ionizing radiation. Additionally, previous approaches in the literature are often time-consuming, making them unrealistic for widespread use in diverse settings, particularly in rural areas. In addition to this, the traditional models struggle to predict highest risk plaques that cause a stroke. To overcome these challenges, the proposed model introduces an innovative approach for classifying carotid plaques using deep learning techniques. In this proposed model, the required ultrasound images are gathered from the available dataset. Further the gathered images are input to the segmentation phase, where the segmentation is carried out using the designed Mask Region-based Convolutional Neural Network (MRCNN) technique. After segmentation, the classification of carotid plaque is performed using the Gabor Convolutional Neural Network with Long Short-Term Memory (GCNN-LSTM). In the end, the experiments are conducted for the developed mechanism for proving its effectiveness over traditional models.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Novel Classification Mechanism of Carotid Plaque Disease Using Gabor CNN with LSTM Layer

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Naga Prudhvi Raj Vattikuti,
  • Samiappan Dhanalakshmi

摘要

Carotid Artery Stenosis (CAS) is the condition, where the carotid artery blocks the blood supply to the brain. This can lead to strokes and heart attacks. In general, the CAS caused by the atherosclerosis, where in the arteries build the plaque and block the blood flow. Medical imaging techniques are used to detect the CAS; however, it often faces limitations like high costs and challenges in exposure to ionizing radiation. Additionally, previous approaches in the literature are often time-consuming, making them unrealistic for widespread use in diverse settings, particularly in rural areas. In addition to this, the traditional models struggle to predict highest risk plaques that cause a stroke. To overcome these challenges, the proposed model introduces an innovative approach for classifying carotid plaques using deep learning techniques. In this proposed model, the required ultrasound images are gathered from the available dataset. Further the gathered images are input to the segmentation phase, where the segmentation is carried out using the designed Mask Region-based Convolutional Neural Network (MRCNN) technique. After segmentation, the classification of carotid plaque is performed using the Gabor Convolutional Neural Network with Long Short-Term Memory (GCNN-LSTM). In the end, the experiments are conducted for the developed mechanism for proving its effectiveness over traditional models.