Early Lung Cancer Detection From CT Scans Using Multi-Scale Image Segmentation

Author(s)	Mr. Vinay Vastrakar, Ms. Nidhi Sharma
Country	India
Abstract	This paper reviews four generations of pulmonary nodule segmentation techniques — classical image processing, patch-based CNN detection, encoder-decoder networks, and multi-scale ASPP-based architectures — and presents a reproducible case study using a MobileNetV2 encoder, ASPP bottleneck, and U-Net decoder trained on the LUNA16 dataset (subset0). The model achieved a test Dice score of 0.595 and IoU of 0.438, reported transparently from logged training results along with identified limitations including patient-level data split risk and an incomplete ASPP ablation.
Keywords	Lung cancer detection, medical image segmentation, convolutional neural networks, U-Net, atrous spatial pyramid pooling, multi-scale learning, transfer learning, LUNA16, computer-aided diagnosis
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 8, Issue 4, July-August 2026
Published On	2026-08-22

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research