IIUM Repository

The classification of lung cancer: a DenseNet feature-based transfer learning evaluation

Arzmi, Mohd Hafiz and P.P. Abdul Majeed, Anwar and Musa, Rabiu Muazu and Mohd Razman, Mohd Azraai and Gan, Hong-Seng and Mohd Khairuddin, Ismail and Ab. Nasir, Ahmad Fakhri (2023) The classification of lung cancer: a DenseNet feature-based transfer learning evaluation. In: Deep Learning in Cancer Diagnostics. SpringerBriefs in Applied Sciences and Technology (1). Springer, Singapore, pp. 21-26. ISBN 978-981-19-8936-0

[img] PDF (Book Chapter) - Published Version
Restricted to Repository staff only

Download (289kB) | Request a copy


In the present study, a class of deep learning approaches is used to clas- sify non-small-cell lung cancers. A total of 400 computed tomography (CT) images of lung cancer that are demarcated into normal, large cell carcinoma, adenocarci- noma and squamous cell carcinoma are split into the 70:15:15 ratio for training, testing and validation. The images are evaluated on different DenseNet-Support Vector Machine (SVM) pipelines, i.e., DenseNet121-SVM, DenseNet169-SVM and DenseNet201-SVM, respectively. It was shown from the present investigation that the DenseNet121-SVM pipeline is able to yield a test classification accuracy of 87%. Therefore, it could be demonstrated that the proposed architecture is able to classify the different variations of non-small-cell lung cancers reasonably well and could further facilitate the diagnosis of lung cancer by clinicians.

Item Type: Book Chapter
Uncontrolled Keywords: Computer-Aided diagnosis; Transfer learning; Lung cancer
Subjects: R Medicine > R Medicine (General)
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button): Kulliyyah of Dentistry > Department of Fundamental Dental and Medical Sciences
Kulliyyah of Dentistry
Depositing User: AP Ts Dr Mohd Hafiz Arzmi
Date Deposited: 09 Mar 2023 14:14
Last Modified: 09 Mar 2023 14:14
URI: http://irep.iium.edu.my/id/eprint/103898

Actions (login required)

View Item View Item


Downloads per month over past year