Scholarlink

<p>With the advancement of the Internet of Things (IoT), tremendous progress has been witnessed in the medical domain. The utilization of Artificial Intelligence (AI) in this domain has been found to assist medical professionals in making more informed decisions, considering the treatment course of patients. Lung cancer is one of the deadliest forms, resulting in a high mortality rate. Generally, lung cancer is detected using Computed Tomography (CT) scans, but with the rising number of CT scans, it is necessary to develop effective and automated methods of detecting lung cancer. This work presents a new lung cancer detection scheme and a multi-objective-based optimization for data transfer from nodes to the Base Station (BS). Here, a hybrid algorithmic approach named Walruses Political Optimization (WPO) is formulated for routing the CT images to the BS based on various objectives. At BS, the CT images undergo various processes, and finally, lung cancer is detected using the Deep Maxout Network (DMN), where the task is formulated as a patient-level diagnosis, classifying each patient as cancerous or non-cancerous based on CT scan data, and the WPO is applied to tune the DMN. Evaluation of the WPO for routing shows that residual energy of 0.030&#xa0;J and delay of 1.936&#xa0;ms are attained. The residual energy obtained by the WPO is 36.67%, 20%, 3.33%, and 10% higher than the residual energy of ESEERP, CAR-MOSOA, HOA-IoT-WSN, and WOA-SA, respectively. Similarly, the WPO-DMN is found to attain accuracy, sensitivity, and specificity of 0.939, 0.950, and 0.947, respectively. Here, the accuracy of the WPO-DMN is 3.83%, 3.41%, 2.45%, and 2.02% higher than the accuracy of ML‑CNN, SLD-SC, HTARDTC, and DL with NL Bayes, respectively.</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

IoT-Based Lung Cancer Detection Using Deep Maxout Network and Walruses Political Optimization-Based Routing Algorithm

  • 期刊论文

  • 发表日期：

    2026年7月20日

查看全文
  • Janardan A Pawar,
  • Ashvini Pradeep Shende,
  • Manoj Ashok Wakchaure,
  • Amol Dhumane,
  • Mubin Tamboli

摘要

With the advancement of the Internet of Things (IoT), tremendous progress has been witnessed in the medical domain. The utilization of Artificial Intelligence (AI) in this domain has been found to assist medical professionals in making more informed decisions, considering the treatment course of patients. Lung cancer is one of the deadliest forms, resulting in a high mortality rate. Generally, lung cancer is detected using Computed Tomography (CT) scans, but with the rising number of CT scans, it is necessary to develop effective and automated methods of detecting lung cancer. This work presents a new lung cancer detection scheme and a multi-objective-based optimization for data transfer from nodes to the Base Station (BS). Here, a hybrid algorithmic approach named Walruses Political Optimization (WPO) is formulated for routing the CT images to the BS based on various objectives. At BS, the CT images undergo various processes, and finally, lung cancer is detected using the Deep Maxout Network (DMN), where the task is formulated as a patient-level diagnosis, classifying each patient as cancerous or non-cancerous based on CT scan data, and the WPO is applied to tune the DMN. Evaluation of the WPO for routing shows that residual energy of 0.030 J and delay of 1.936 ms are attained. The residual energy obtained by the WPO is 36.67%, 20%, 3.33%, and 10% higher than the residual energy of ESEERP, CAR-MOSOA, HOA-IoT-WSN, and WOA-SA, respectively. Similarly, the WPO-DMN is found to attain accuracy, sensitivity, and specificity of 0.939, 0.950, and 0.947, respectively. Here, the accuracy of the WPO-DMN is 3.83%, 3.41%, 2.45%, and 2.02% higher than the accuracy of ML‑CNN, SLD-SC, HTARDTC, and DL with NL Bayes, respectively.