Scholarlink

Purpose <p>As lung carcinoma is a fatal disease, accurate classification from Histopathological images of the tumor is a challenging task. The shapes of the tumors are irregular, and there are imaging differences. The objective of this study was to create a reliable ensemble deep learning technique for the detection and classification of types of lung cancer.</p> Methods <p>In this paper, an Advanced Ensemble Deep Learning Approach for Accurate and Reliable Classification of Histopathological Image Classification in Lung Cancer Diagnosis (EDL-ARC-HIC-LCD) is proposed. Initially, input image is collected from Lung Cancer (Histopathological Images) dataset. Then pre-processed using Non-Uniform Weighted Guided Filtering (NWGF) is used for data normalization and noise reduction. Afterwards the pre-processed images are given to feature extraction and Lung Cancer Detection using EfficientNetB3, MobileNetV1 and DenseNet121 is used detect lung cancer and classify as Squamous Cell Carcinoma, Adenocarcinoma, Benign. Finally, Groupers and Moray Eels Optimization (GMEO) is employed to optimize EfficientNetB3, MobileNetV1 and DenseNet121 for precisely detect and classify in lung cancer. The outputs from EfficientNetB3, MobileNetV1 and DenseNet121 are combined using a stacking ensemble strategy.</p> Results <p>This sophisticated technique enhances classification accuracy by refining the predictions from each model.The proposed EDL-ARC-HIC-LCD attains26.68%, 25.75%, and 26.16% better accuracy and 27.49%, 24.75%, and 25.85% higher precision when compared with existing methods.</p> Conclusion <p>The EDL-ARC-HIC-LCD method is effective in enhancing the detection and classification of lung cancer, providing a reliable tool in the automated diagnosis of the disease.</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

An advanced ensemble deep learning approach for accurate and reliable classification of histopathological image classification in lung cancer diagnosis

  • 期刊论文

  • 发表日期：

    2026年7月6日

查看全文
  • R. Renugadevi Rajaram,
  • Midhun Chakkaravarthy J,
  • Janani Selvam,
  • Ashok Vajravelu

摘要

Purpose

As lung carcinoma is a fatal disease, accurate classification from Histopathological images of the tumor is a challenging task. The shapes of the tumors are irregular, and there are imaging differences. The objective of this study was to create a reliable ensemble deep learning technique for the detection and classification of types of lung cancer.

Methods

In this paper, an Advanced Ensemble Deep Learning Approach for Accurate and Reliable Classification of Histopathological Image Classification in Lung Cancer Diagnosis (EDL-ARC-HIC-LCD) is proposed. Initially, input image is collected from Lung Cancer (Histopathological Images) dataset. Then pre-processed using Non-Uniform Weighted Guided Filtering (NWGF) is used for data normalization and noise reduction. Afterwards the pre-processed images are given to feature extraction and Lung Cancer Detection using EfficientNetB3, MobileNetV1 and DenseNet121 is used detect lung cancer and classify as Squamous Cell Carcinoma, Adenocarcinoma, Benign. Finally, Groupers and Moray Eels Optimization (GMEO) is employed to optimize EfficientNetB3, MobileNetV1 and DenseNet121 for precisely detect and classify in lung cancer. The outputs from EfficientNetB3, MobileNetV1 and DenseNet121 are combined using a stacking ensemble strategy.

Results

This sophisticated technique enhances classification accuracy by refining the predictions from each model.The proposed EDL-ARC-HIC-LCD attains26.68%, 25.75%, and 26.16% better accuracy and 27.49%, 24.75%, and 25.85% higher precision when compared with existing methods.

Conclusion

The EDL-ARC-HIC-LCD method is effective in enhancing the detection and classification of lung cancer, providing a reliable tool in the automated diagnosis of the disease.