使用基于生物启发的深度学习模型增强超调谐,用于准确的肺癌检测和分类。

IF 1.3 4区 医学 Q4 ENGINEERING, BIOMEDICAL
Jyoti Kumari, Sapna Sinha, Laxman Singh
{"title":"使用基于生物启发的深度学习模型增强超调谐,用于准确的肺癌检测和分类。","authors":"Jyoti Kumari, Sapna Sinha, Laxman Singh","doi":"10.1177/03913988251359522","DOIUrl":null,"url":null,"abstract":"<p><p>Lung cancer (LC) is one of the leading causes of cancer related deaths worldwide and early recognition is critical for enhancing patient outcomes. However, existing LC detection techniques face challenges such as high computational demands, complex data integration, scalability limitations, and difficulties in achieving rigorous clinical validation. This research proposes an Enhanced Hyper Tuning Deep Learning (EHTDL) model utilizing bioinspired algorithms to overcome these limitations and improve accuracy and efficiency of LC detection and classification. The methodology begins with the Smooth Edge Enhancement (SEE) technique for preprocessing CT images, followed by feature extraction using GLCM-based Texture Analysis. To refine the features and reduce dimensionality, a Hybrid Feature Selection approach combining Grey Wolf optimization (GWO) and Differential Evolution (DE) is employed. Precise lung segmentation is performed using Mask R-CNN to ensure accurate delineation of lung regions. A Deep Fractal Edge Classifier (DFEC) is introduced, consisting of five fractal blocks with convolutional layers and pooling to progressively learn LC characteristics. The proposed EHTDL model achieves remarkable performance metrics, including 99% accuracy, 100% precision, 98% recall, and 99% <i>F</i>1-score, demonstrating its robustness and effectiveness. The model's scalability and efficiency make it suitable for real-time clinical application offering a promising solution for early LC detection and significantly enhancing patient care.</p>","PeriodicalId":13932,"journal":{"name":"International Journal of Artificial Organs","volume":" ","pages":"3913988251359522"},"PeriodicalIF":1.3000,"publicationDate":"2025-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Enhanced hyper tuning using bioinspired-based deep learning model for accurate lung cancer detection and classification.\",\"authors\":\"Jyoti Kumari, Sapna Sinha, Laxman Singh\",\"doi\":\"10.1177/03913988251359522\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>Lung cancer (LC) is one of the leading causes of cancer related deaths worldwide and early recognition is critical for enhancing patient outcomes. However, existing LC detection techniques face challenges such as high computational demands, complex data integration, scalability limitations, and difficulties in achieving rigorous clinical validation. This research proposes an Enhanced Hyper Tuning Deep Learning (EHTDL) model utilizing bioinspired algorithms to overcome these limitations and improve accuracy and efficiency of LC detection and classification. The methodology begins with the Smooth Edge Enhancement (SEE) technique for preprocessing CT images, followed by feature extraction using GLCM-based Texture Analysis. To refine the features and reduce dimensionality, a Hybrid Feature Selection approach combining Grey Wolf optimization (GWO) and Differential Evolution (DE) is employed. Precise lung segmentation is performed using Mask R-CNN to ensure accurate delineation of lung regions. A Deep Fractal Edge Classifier (DFEC) is introduced, consisting of five fractal blocks with convolutional layers and pooling to progressively learn LC characteristics. The proposed EHTDL model achieves remarkable performance metrics, including 99% accuracy, 100% precision, 98% recall, and 99% <i>F</i>1-score, demonstrating its robustness and effectiveness. The model's scalability and efficiency make it suitable for real-time clinical application offering a promising solution for early LC detection and significantly enhancing patient care.</p>\",\"PeriodicalId\":13932,\"journal\":{\"name\":\"International Journal of Artificial Organs\",\"volume\":\" \",\"pages\":\"3913988251359522\"},\"PeriodicalIF\":1.3000,\"publicationDate\":\"2025-08-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Journal of Artificial Organs\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.1177/03913988251359522\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, BIOMEDICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Artificial Organs","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1177/03913988251359522","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}
引用次数: 0

摘要

肺癌(LC)是全球癌症相关死亡的主要原因之一,早期识别对于提高患者预后至关重要。然而,现有的LC检测技术面临着诸如高计算需求、复杂的数据集成、可扩展性限制以及难以实现严格的临床验证等挑战。本研究提出了一种利用生物启发算法的增强型超调优深度学习(ehtml)模型,以克服这些限制,提高LC检测和分类的准确性和效率。该方法首先使用平滑边缘增强(SEE)技术对CT图像进行预处理,然后使用基于glcm的纹理分析进行特征提取。采用灰狼优化(GWO)和差分进化(DE)相结合的混合特征选择方法对特征进行细化和降维。使用Mask R-CNN进行精确的肺分割,以确保肺区域的准确描绘。介绍了一种深度分形边缘分类器(Deep Fractal Edge Classifier, DFEC),它由五个具有卷积层的分形块和池化组成,逐步学习LC特征。提出的ehtml模型实现了显著的性能指标,包括99%的准确率、100%的精度、98%的召回率和99%的f1分数,证明了其鲁棒性和有效性。该模型的可扩展性和高效性使其适合于实时临床应用,为早期LC检测提供了有前途的解决方案,并显著提高了患者护理。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Enhanced hyper tuning using bioinspired-based deep learning model for accurate lung cancer detection and classification.

Lung cancer (LC) is one of the leading causes of cancer related deaths worldwide and early recognition is critical for enhancing patient outcomes. However, existing LC detection techniques face challenges such as high computational demands, complex data integration, scalability limitations, and difficulties in achieving rigorous clinical validation. This research proposes an Enhanced Hyper Tuning Deep Learning (EHTDL) model utilizing bioinspired algorithms to overcome these limitations and improve accuracy and efficiency of LC detection and classification. The methodology begins with the Smooth Edge Enhancement (SEE) technique for preprocessing CT images, followed by feature extraction using GLCM-based Texture Analysis. To refine the features and reduce dimensionality, a Hybrid Feature Selection approach combining Grey Wolf optimization (GWO) and Differential Evolution (DE) is employed. Precise lung segmentation is performed using Mask R-CNN to ensure accurate delineation of lung regions. A Deep Fractal Edge Classifier (DFEC) is introduced, consisting of five fractal blocks with convolutional layers and pooling to progressively learn LC characteristics. The proposed EHTDL model achieves remarkable performance metrics, including 99% accuracy, 100% precision, 98% recall, and 99% F1-score, demonstrating its robustness and effectiveness. The model's scalability and efficiency make it suitable for real-time clinical application offering a promising solution for early LC detection and significantly enhancing patient care.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
International Journal of Artificial Organs
International Journal of Artificial Organs 医学-工程:生物医学
CiteScore
3.40
自引率
5.90%
发文量
92
审稿时长
3 months
期刊介绍: The International Journal of Artificial Organs (IJAO) publishes peer-reviewed research and clinical, experimental and theoretical, contributions to the field of artificial, bioartificial and tissue-engineered organs. The mission of the IJAO is to foster the development and optimization of artificial, bioartificial and tissue-engineered organs, for implantation or use in procedures, to treat functional deficits of all human tissues and organs.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信