Lung Cancer Detection and Classification using Deep Learning

Lung Cancer Detection and Classification using Deep Learning

IEEE BASE PAPER TITLE:

Lung-AttNet: An Attention Mechanism-Based CNN Architecture for Lung Cancer Detection With Federated Learning

IEEE BASE PAPER ABSTRACT:

Lung cancer is one of the fatal diseases whose early diagnosis is essential to mitigate the death rate. Computed Tomography (CT) scans are widely used for lung cancer diagnosis, but manual interpretation by health professionals can lead to inconsistent results. To address this, we propose Lung-AttNet, a novel lightweight convolutional neural network (CNN) model enhanced with an attention mechanism. Lung-AttNet incorporates a convolutional block with a Lightweight Global Attention Module (LGAM) to effectively distinguish between lung cancer types. The convolutional block extracts both low- and high-dimensional features, while LGAM captures feature dependencies across channel and spatial dimensions. The model is evaluated using the Kaggle CT scan dataset, which includes four classes: adenocarcinoma, large cell carcinoma, squamous cell carcinoma, and normal. Extensive experiments, including ablation studies, 5-fold cross-validation, and explainable AI (XAI) techniques such as Grad-CAM and LIME, demonstrate that Lung-AttNet achieves an average accuracy of 91.5%. Furthermore, to address medical data sensitivity and privacy concerns, the model is deployed in a Federated Learning (FL) framework, where the global model is trained using weights from local models rather than sharing raw data. In the FL environment, Lung-AttNet achieves an accuracy of 92% with 2 and 3 clients, underscoring its robustness and adaptability for real-world applications.

PROJECT OUTPUT VIDEO:

ALGORITHM /MODEL USED:

Convolutional Neural Network (CNN) based on the VGG16 architecture.

OUR PROPOSED ABSTRACT:

Lung cancer remains one of the leading causes of cancer-related mortality worldwide, primarily due to late-stage diagnosis and the subtle nature of early symptoms. Advances in medical imaging and deep learning have opened new possibilities for automated, accurate, and early detection of lung cancer from CT scan images. In this context, the project “Lung Cancer Detection and Classification using Deep Learning” focuses on building an intelligent, web-enabled diagnostic support system that can assist clinicians in identifying and classifying lung cancer types with high reliability.

The need for this system arises from the growing volume of radiological data and the dependency on manual interpretation, which is time-consuming and prone to inter-observer variability. Automated deep learning-based solutions can significantly reduce diagnostic effort while improving consistency and accuracy. By leveraging convolutional neural networks, the project aims to enhance early detection, support multi-class classification of lung cancer, and provide clear performance insights through visual analytics.

The developed system is implemented using Python as the core programming language, with HTML, CSS, and JavaScript for the front end, and the Flask framework for integrating the model into a user-friendly web application. The backend employs a deep convolutional neural network based on the VGG16 architecture. To improve feature representation beyond the standard VGG16 model, residual connections and Squeeze-and-Excitation (SE) blocks are integrated, enabling better gradient flow and channel-wise feature recalibration.

The dataset used in this project consists of 1,000 CT scan chest images distributed across four classes: Adenocarcinoma (338 images), Large Cell Carcinoma (187 images), Squamous Cell Carcinoma (260 images), and Normal cases (215 images). The proposed model achieved a training accuracy of 99.67% and a test accuracy of 95.24%, demonstrating strong generalization performance. Comprehensive performance evaluation is carried out using metrics such as accuracy, precision, recall, specificity, F1-score, and a confusion matrix to analyze class-wise prediction behavior.

In addition to quantitative evaluation, the system presents multiple visualization charts to enhance interpretability and analysis. These include dataset class distribution charts, model accuracy comparison charts, training versus validation loss curves, and prediction confidence distribution charts. Together, these components form a complete, robust, and interpretable deep learning-based lung cancer detection and classification system that effectively bridges advanced AI techniques with practical clinical decision support.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS: 

• System : Pentium i3 Processor.
• Hard Disk : 20 GB.
• Monitor : 15’’ LED.
• Input Devices : Keyboard, Mouse.
• Ram : 8 GB.

SOFTWARE REQUIREMENTS: 

• Operating System : Windows 10 / 11.
• Coding Language : Python 3.11.6.
• Web Framework : Flask.
• Frontend : HTML, CSS, JavaScript.

REFERENCE:

CHAMAK SAHA, SOMAK SAHA, MD. ASADUR RAHMAN, MD. MAHMUDUL HAQUE MILU, HIROKI HIGA, MOHD ABDUR RASHID, AND NASIM AHMED, “Lung-AttNet: An Attention Mechanism-Based CNN Architecture for Lung Cancer Detection With Federated Learning”, IEEE ACCESS, VOLUME 13, 2025.

👉CLICK HERE TO BUY THIS PROJECT “Lung Cancer Detection and Classification using Deep Learning” SOURCE CODE👈

Frequently Asked Questions (FAQ’s) and Answers

Q1. What is the objective of the Lung Cancer Detection and Classification using Deep Learning project?

The main objective of this project is to develop an automated system that can detect and classify lung cancer from CT scan chest images using deep learning techniques. The system aims to assist in identifying lung cancer types accurately and efficiently by analyzing medical images through a trained convolutional neural network.

Q2. Which lung conditions are classified in this project?

The system classifies CT scan images into four categories: Adenocarcinoma, Large Cell Carcinoma, Squamous Cell Carcinoma, and Normal cases. This multi-class classification provides detailed diagnostic insights rather than a simple cancer–non-cancer decision.

Q3. What deep learning model is used in this project?

The project uses a Convolutional Neural Network based on the VGG16 architecture. The base model is enhanced with residual connections and Squeeze-and-Excitation blocks to improve feature extraction and classification performance.

Q4. What programming languages and technologies are used?

The system is developed using Python as the core programming language. The frontend is built using HTML, CSS, and JavaScript, and the backend deployment is handled using the Flask framework.

Q5. What type of dataset is used for training and testing?

The dataset consists of 1,000 CT scan chest images distributed across four classes: Adenocarcinoma, Large Cell Carcinoma, Squamous Cell Carcinoma, and Normal. These images are used for both training and evaluating the performance of the model.

Q6. How accurate is the proposed system?

The proposed system achieves high performance, with a training accuracy of 99.67% and a test accuracy of 95.24%. These results indicate strong learning capability and effective generalization to unseen CT scan images.

Q7. What performance metrics are used to evaluate the model?

The model is evaluated using multiple performance metrics, including accuracy, precision, recall, specificity, F1-score, and confusion matrix. These metrics provide a comprehensive assessment of the system’s classification performance.

Q8. Does the system provide any visual analysis or charts?

Yes, the system generates several visualization charts such as dataset class distribution, model accuracy comparison, training versus validation loss curves, and prediction confidence distribution charts to support better analysis and interpretation of results.

Q9. Does the project require specialized hardware?

No, the project can be developed and deployed using standard computing hardware. Training may benefit from GPU support, but inference and web deployment can run on typical systems with adequate processing power.

Q10. What makes this project suitable for academic and research use?

The project combines deep learning, medical image analysis, performance evaluation, and web-based deployment in a single framework. Its structured methodology, measurable results, and clear visualization outputs make it well-suited for academic learning, research exploration, and project-based evaluation.