EfficientNetB1 Model for Lung Cancer Detection using Biopsy Images
Jess John, Aaradhya Deotale, Figo Fernandez, Dipti Jadhav · 2024
Lung cancer, a pervasive and life-threatening ailment, necessitates early and precise diagnosis. This research endeavors to employ Convolutional Neural Networks (CNNs) to automate the detection of lung cancer in histopathological biopsy images, addressing challenges associated with timely and accurate diagnosis. Given the error-prone and time-consuming nature of manual assessment by pathologists, an automated approach becomes imperative. The study encompasses a comprehensive scope, including enhancing interpretability, classifying specific lung cancer subtypes, real-time intraoperative analysis, and extending the application to other cancer types. The methodology involves the utilization of a pre-trained Efficient Net-based model for image classification, showcasing its efficacy in discerning between benign and malignant lung cancer cells, as evidenced by robust training results. Moreover, the model exhibits potential for personalization in lung cancer diagnosis and treatment. Research findings affirm that machine learning models, specifically the EfficientNet-based architecture, markedly improve lung cancer detection accuracy. The model's proficiency in subtype differentiation and its capacity for real-time surgical analysis represent significant strides in lung cancer diagnostics and treatment. In conclusion, this project addresses the critical imperative for accurate and timely lung cancer diagnosis, providing a promising advancement in combatting this devastating disease. The developed model holds transformative potential in radiology and oncology, serving as a valuable tool for medical professionals and contributing to enhanced patient outcomes.