• Version
• Download 9
• File Size 348.75 KB
• File Count 1
• Create Date 7 June 2023
• Last Updated 7 June 2023

Lungs Cancer Detection and Classification with Image Processing

Ashvani Kumar

Department of Computer Science & Information Technology

Meerut Institute of Engineering & Technology

Fatehpur, Uttar Pradesh, India.

ashvanikumar204@gmail.com

Karan Singh

Department of Computer Science & Information Technology

Meerut Institute of Engineering & Technology

Varanasi, Uttar Pradesh, India.

karansinghrajput5654@gmail.com

Abstract — Lung cancer is a major global health issue, and the timely detection of the disease is essential for enhancing patient outcomes. In recent years, image processing techniques have shown promise in accurately detecting lung cancer. This research paper presents a novel approach for lung cancer detection using image processing, incorporating technologies such as TensorFlow, Tkinter, TFLearn, NumPy, Pandas, Matplotlib, Codecs, OS, Math, and PyIDCOM.

The research methodology begins with data collection, acquiring a dataset of lung images consisting of cancerous and non-cancerous tissues. Preprocessing techniques, including resizing, normalization, noise removal, and histogram equalization, are applied to enhance the image quality and extract relevant features. Advanced feature extraction methods, such as texture analysis, edge detection, and morphological operations, are then employed to identify discriminative characteristics between cancerous and non-cancerous lung tissues.

The extracted features are used to train a lung cancer classification model using TensorFlow, an open-source machine learning framework, with the assistance of TFLearn, a high-level TensorFlow API, to simplify the model creation and training process. To provide an intuitive user interface, Tkinter, a Python library for GUI development, is utilized to create an interactive platform for users to input lung images and obtain real-time predictions regarding the presence of cancerous tissues.

The lung cancer detection system's performance is assessed using different metrics such as accuracy, precision, recall, and F1-score. These metrics are used to compare the predicted labels with the actual labels from an independent test dataset. The findings indicate the efficacy of the proposed system, highlighting its potential for precise and dependable lung cancer detection.

This research contributes to the field of lung cancer detection by combining multiple technologies and providing a user-friendly platform for medical professionals. The system's accuracy and efficiency can aid in early diagnosis, leading to improved patient outcomes and survival rates. Future research directions may involve exploring additional image processing techniques and expanding the dataset for further validation and enhancement of the system.

Keywords: TensorFlow, Tkinter, TFLearn, NumPy, Pandas,