Scholarlink

<p>Lung Disease is a significant contributor to the rising mortality rate, as it impacts overall respiratory well-being. None of the current studies have concentrated on the Severity Level of LD using Spirometric Data, Arterial Blood Gas easurements, and Demographic Risk Factors. Consequently, the study introduces Adaptive Neuro Lipschitz Batch Normalization Fuzzy Inference Systems (ANLipBNFIS) for SL identification derived from Pulmonary Function Test (PFT). Initially, the Chest XRay Image (CHXRI) undergoes pre-processing, and data is balanced through a Generative Adversarial Network (GAN). The nodules and opacities are segmented by Attention Gates-based SegNet (AG-SegNet). Then, the Directional and Radial Patterns (DRP) are recognized using Composite 2 Dimension based Fast Fourier Transform (C2D-FFT). Meanwhile, from the balanced data, the tissues are clustered by K-means Clustering to determine the tissue density variation. Then, the tissue-clustered output is subjected to Topographical Elevation (TE). After extracting the essential features from the segmented image, clustered image, DRP analyzed image, and TE image, the best features are selected using Logistic Sine Chimp Optimization (LS-ChO). Further, LD classification is done by Bias Bend Soft-Root-Sign Visual Geometry Group 16-layer network (B<sup>2</sup>SRS-VGG16). At last, SL is identified by employing ANLipBNFIS. Hence, the proposed model identified the severity of LD within a fuzzification time of 2589&#xa0;ms, showing better performance than the prevailing models.</p>

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

Pulmonary Function Testing Based Severity Level Identification Using ANLipBNFIS and Lung Disease Classification Using B2SRS-VGG16

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月14日

查看全文
  • N. Gobalakrishnan,
  • N. Gobinathan

摘要

Lung Disease is a significant contributor to the rising mortality rate, as it impacts overall respiratory well-being. None of the current studies have concentrated on the Severity Level of LD using Spirometric Data, Arterial Blood Gas easurements, and Demographic Risk Factors. Consequently, the study introduces Adaptive Neuro Lipschitz Batch Normalization Fuzzy Inference Systems (ANLipBNFIS) for SL identification derived from Pulmonary Function Test (PFT). Initially, the Chest XRay Image (CHXRI) undergoes pre-processing, and data is balanced through a Generative Adversarial Network (GAN). The nodules and opacities are segmented by Attention Gates-based SegNet (AG-SegNet). Then, the Directional and Radial Patterns (DRP) are recognized using Composite 2 Dimension based Fast Fourier Transform (C2D-FFT). Meanwhile, from the balanced data, the tissues are clustered by K-means Clustering to determine the tissue density variation. Then, the tissue-clustered output is subjected to Topographical Elevation (TE). After extracting the essential features from the segmented image, clustered image, DRP analyzed image, and TE image, the best features are selected using Logistic Sine Chimp Optimization (LS-ChO). Further, LD classification is done by Bias Bend Soft-Root-Sign Visual Geometry Group 16-layer network (B2SRS-VGG16). At last, SL is identified by employing ANLipBNFIS. Hence, the proposed model identified the severity of LD within a fuzzification time of 2589 ms, showing better performance than the prevailing models.