The early and precise diagnosis of dangerous diseases like cancer considerably improves patient outcomes and treatment effectiveness, making cancer detection and diagnosis essential components of modern healthcare. This chapter explores the development of cancer detection procedures, from early attempts to the most recent advancements produced by deep learning techniques. The urgent need to lessen the severe effects of this disease drives the necessity of cancer detection. The likelihood of a patient’s successful recovery is increased by early discovery, which also makes it possible to pursue more effective and relatively minimal treatment alternatives. Precision in the detection and classification of tumors can revolutionize cancer treatment, lowering mortality rates and enhancing the quality of life for cancer patients. Manual assessments and rule-based algorithms were once the foundations of cancer detection techniques, but they frequently lacked the precise discriminations needed for different types of tumors. Machine learning approaches led to small advancements, but deep learning brought about the breakthrough. Due to their proficiency in automated feature extraction and pattern recognition, convolutional neural networks (CNNs) have revolutionized cancer detection using a variety of modalities. Notable examples include “Cancer diagnosis in histopathological image: CNN based approach” (Elhaj et al. Sudan Med Monit 10:1, 2015) and “Deep Learning Predicts Lung Cancer Treatment Response from Serial Medical Imaging” (Alsuliman et al., Curr Res Trans Med 68:245–251, 2020). The field experienced improved model generalization and adaptability with the introduction of transfer learning, opening the door for the fine-tuning of pretrained models for particular cancer types. This development sped up the transition to multimodal fusion approaches, utilizing data from several imaging sources for thorough cancer assessment. One such example is “A Multimodal Deep Neural Network for Human Breast Cancer Prognosis Prediction by Integrating Multi-Dimensional Data” (Bhattacharyya et al., Deep learning research with engineering applications. De Gruyter Publications, 2020). This chapter aims to provide a thorough analysis of the development of cancer detection techniques, outlining each technique’s advantages, disadvantages, and practical applications. This work contributes to the expanding body of knowledge by providing insights into the development of cancer detection, particularly through the lens of cutting-edge deep learning techniques. These insights are crucial for providing medical professionals with the tools they need to diagnose cancer early and accurately.

Deep Learning-Based Techniques for Detection of Cancer

章节
发表日期：

2025年

查看全文

Dhruv Umesh Sompura,
B. K. Tripathy

摘要

The early and precise diagnosis of dangerous diseases like cancer considerably improves patient outcomes and treatment effectiveness, making cancer detection and diagnosis essential components of modern healthcare. This chapter explores the development of cancer detection procedures, from early attempts to the most recent advancements produced by deep learning techniques. The urgent need to lessen the severe effects of this disease drives the necessity of cancer detection. The likelihood of a patient’s successful recovery is increased by early discovery, which also makes it possible to pursue more effective and relatively minimal treatment alternatives. Precision in the detection and classification of tumors can revolutionize cancer treatment, lowering mortality rates and enhancing the quality of life for cancer patients. Manual assessments and rule-based algorithms were once the foundations of cancer detection techniques, but they frequently lacked the precise discriminations needed for different types of tumors. Machine learning approaches led to small advancements, but deep learning brought about the breakthrough. Due to their proficiency in automated feature extraction and pattern recognition, convolutional neural networks (CNNs) have revolutionized cancer detection using a variety of modalities. Notable examples include “Cancer diagnosis in histopathological image: CNN based approach” (Elhaj et al. Sudan Med Monit 10:1, 2015) and “Deep Learning Predicts Lung Cancer Treatment Response from Serial Medical Imaging” (Alsuliman et al., Curr Res Trans Med 68:245–251, 2020). The field experienced improved model generalization and adaptability with the introduction of transfer learning, opening the door for the fine-tuning of pretrained models for particular cancer types. This development sped up the transition to multimodal fusion approaches, utilizing data from several imaging sources for thorough cancer assessment. One such example is “A Multimodal Deep Neural Network for Human Breast Cancer Prognosis Prediction by Integrating Multi-Dimensional Data” (Bhattacharyya et al., Deep learning research with engineering applications. De Gruyter Publications, 2020). This chapter aims to provide a thorough analysis of the development of cancer detection techniques, outlining each technique’s advantages, disadvantages, and practical applications. This work contributes to the expanding body of knowledge by providing insights into the development of cancer detection, particularly through the lens of cutting-edge deep learning techniques. These insights are crucial for providing medical professionals with the tools they need to diagnose cancer early and accurately.