A Novel Method to Detect Lymphangioleiomyomatosis (LAM) Using CNNLSTM on Computed Tomography
Sithika Seema. S1, Sumathy. R2
1Sithika Seema. S, Department of Computer Science and Engineering, Kalaignarkarunanidhi Institute of Technology Coimbatore (Tamil Nadu), India.
2Ms. Sumathy. R, Assistant Professor, Kalaignarkarunanidhi Institute of Technology Coimbatore (Tamil Nadu), India.
Manuscript received on 29 October 2024 | Revised Manuscript received on 05 November 2024 | Manuscript Accepted on 15 November 2024 | Manuscript published on 30 November 2024 | PP: 24-28 | Volume-5 Issue-1, November 2024 | Retrieval Number: 100.1/ijpmh.B105205020125 | DOI: 10.54105/ijpmh.B1052.05011124
Open Access | Ethics and Policies | Cite | Zenodo | OJS | Indexing and Abstracting
© The Authors. Published by Lattice Science Publication (LSP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: “Detection of Lymphangioleiomyomatosis (LAM) using CNN-LSTM framework on CT images” Lymphangioleiomyomatosis (LAM) is a progressive lung disease marked by uncommon cystic growths, often diagnosed using high-resolution CT scan images. The complex nature of LAM’s radiological findings and the need for specialised knowledge make an early and accurate diagnosis difficult, even though it is essential for successful treatment. To automate the detection of LAM in CT images, this study introduces an advanced deep learning approach that integrates a CNN with an LSTM network. A well-annotated dataset of CT scans from patients with LAM and healthy controls was used to create and evaluate the model. To increase model robustness and generalizability, extensive preprocessing was used, including lung area segmentation and data augmentation. Accuracy, sensitivity, specificity, and AUCROC were among the key performance indicators that demonstrated the model’s ability to distinguish between LAM and non-LAM cases. The accuracy of the CNN-LSTM model for detecting Lymphangioleiomyomatosis (LAM) using CT scan images is approximately 80%. By providing a dependable, noninvasive, and scalable approach to early LAM identification, this novel CNN-LSTM design lessens the need for expert interpretation and improves diagnostic effectiveness.
Keywords: Lymphangioleiomyomatosis (LAM) disease, Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Deep Learning, Image Classification, CT Scan Images.
Scope of the Article: Community Health