Influence of Input Image Size Variations and Data Balancing on VGG-16 and VGG-16-ELM Models  for Pneumonia Classification

Mohammad Agil Rofiqul Zein; Basuki  Rahmat; Eva Yulia Puspaningrum

doi:10.59934/jaiea.v4i3.1166

Authors

Mohammad Agil Rofiqul Zein Universitas Pembangunan Nasional Veteran Jawa Timur
Basuki Rahmat Universitas Pembangunan Nasional Veteran Jawa Timur
Eva Yulia Puspaningrum Universitas Pembangunan Nasional Veteran Jawa Timur

DOI:

https://doi.org/10.59934/jaiea.v4i3.1166

Keywords:

Pneumonia, Deep Learning, VGG-16, Extreme Learning Machine, Data Balancing

Abstract

Pneumonia is a lung disease that can be identified through chest X-ray images. This study aims to evaluate the performance of two deep learning models, namely VGG-16 and a combination of VGG-16 with Extreme Learning Machine (ELM), in automatically classifying pneumonia. The approach used includes an analysis of variations in input image sizes (150×150, 200×200, 224×224, 256×256, and 300×300 pixels) as well as the application of data balancing techniques using Random Over Sampling (ROS). The dataset used contains 5,856 X-ray images classified into two classes: NORMAL and PNEUMONIA. The preprocessing stages include resizing, normalization, data splitting, and augmentation. Performance evaluation is conducted using metrics of accuracy, precision, recall, and F1-score. The experimental results show that the input size of 200×200 consistently yields the best performance. The VGG-16 model without the application of ROS achieved the highest accuracy of 96.59% and an F1-score of 97.69%. Meanwhile, the VGG-16-ELM combination showed significant performance improvement when ROS was applied. These findings indicate that the selection of model architecture, data balancing techniques, and input image size significantly influence classification accuracy, and contribute to the development of AI-based automated diagnostic systems.

Keywords: Pneumonia, Deep Learning, VGG-16, Extreme Learning Machine, Data Balancing.

Downloads

Download data is not yet available.

References

F. Antony, H. Irsyad, and M. E. Al Rivan, "KNN and Gabor Filter as well as Wiener Filter for Diagnosing Pneumonia in X-RAY Images of the Lungs," J. Algoritm., vol. 1, no. 2, pp. 147–155, 2021, doi: 10.35957/algoritme.v1i2.893.

R. A. Wati, H. Irsyad, and M. E. A. R. Rivan, "Pneumonia Classification Using Support Vector Machine Method," J. Algoritm., vol. 1, no. 1, pp. 21–32, 2020, [Online]. Available: http://download.garuda.kemdikbud.go.id/article.php?article=2652619&val=24585&title=PNEUMONIA CLASSIFICATION USING SUPPORT VECTOR MACHINE.

T. Berliani, E. Rahardja, and L. Septiana, "Comparison of X-ray Lung Image Classification Ability Using Transfer Learning ResNet-50 and VGG-16," J. Med. Heal., vol. 5, no. 2, pp. 123–135, 2023, doi: 10.28932/jmh.v5i2.6116.

M. A. Ridho and M. Arman, "Analysis of DDoS Attacks Using Artificial Neural Network Method," J. Sisfokom (Information Systems and Computer), vol. 9, no. 3, pp. 373–379, 2020, doi: 10.32736/sisfokom.v9i3.945.

D. Candra, G. Wibisono, M. Ayu, and M. Afrad, "Transfer Learning Convolutional Neural Network model using VGG-16 for Brain Tumor Classification on MRI Images," LEDGER J. Inform. Inf. Technol., vol. 3, no. 1, pp. 11–18, 2024.

B. Nugroho and E. Y. Puspaningrum, "Performance of CNN Method for Pneumonia Classification with Variations in Input Image Sizes," J. Teknol. Inf. dan Ilmu Komput., vol. 8, no. 3, pp. 533–538, 2021, doi: 10.25126/jtiik.2021834515.

V. V. Nurdiansyah, I. Cholissodin, and P. P. Adikara, "Tuberculosis (TB) Disease Classification using Extreme Learning Machine (ELM) Method," J. Pengemb. Teknol. Inf. dan Ilmu Komput., vol. 4, no. 5, pp. 1387–1393, 2020.

S. A. Suryaman, R. Magdalena, and S. Sa’idah, "Weather Classification Using VGG-16 Method, Principal Component Analysis, and K-Nearest Neighbor," J. Ilmu Komput. dan Inform., vol. 1, no. 1, pp. 1–8, 2021, doi: 10.54082/jiki.1.

M. Harits, R. Ramadhan, and Y. Miftahuddin, “The Effect of Data Balancing on Pneumonia Classification Using VGG16,” pp. 1–11, 2023.

S. N. Nugraha, R. Pebrianto, and E. Fitri, “Application of Deep Learning in the Classification of Paprika Plants Based on Leaf Images Using the CNN Method,” Inf. Syst. Educ. Prof. J. Inf. Syst., vol. 8, no. 2, p. 15, 2023, doi: 10.51211/isbi.v8i2.2671.

P. P. Winangun, I. M. O. Widyantara, and R. S. Hartati, “A Diagnostic Approach Based on Extreme Learning Machine with Linear Kernel to Classify Lung Abnormalities,” Maj. Ilm. Teknol. Elektro, vol. 19, no. 1, p. 83, 2020, doi: 10.24843/mite.2020.v19i01.p12.

L. Listyalina et al., “Determination of Lung Diseases Using Artificial Neural Networks,” J. SIMETRIS, vol. 11, no. 1, pp. 233–240, 2020.

R. AGUSTINA, R. MAGDALENA, and N. K. C. PRATIWI, “Classification of Skin Cancer using Convolutional Neural Network Method with VGG-16 Architecture,” ELKOMIKA J. Technol. Electr. Energy, vol. 10, no. 2, p. 446, 2022, doi: 10.26760/elkomika.v10i2.446.

W. Swastika, “Preliminary Study on COVID-19 Detection Using CT Images Based on Deep Learning,” J. Technol. Inf. and Computer Science, vol. 7, no. 3, pp. 629–634, 2020, doi: 10.25126/jtiik.2020733399.

S. Fadnavis, “Image Interpolation Techniques in Digital Image Processing: An Overview,” J. Eng. Res. Appl. www.ijera.com, vol. 4, no. November 2014, pp. 70–73, 2014, [Online]. Available: www.ijera.com.

R. Mohammed, J. Rawashdeh, and M. Abdullah, “Machine Learning with Oversampling and Undersampling Techniques: Overview Study and Experimental Results,” 2020 11th Int. Conf. Info. Commun. Syst. ICICS 2020, pp. 243–248, 2020, doi: 10.1109/ICICS49469.2020.239556.

World Health Organization, 'Air pollution and health,' 2021.

Ministry of Health of the Republic of Indonesia, 'Pneumonia Continues to Threaten Children,' Ministry of Health RI, November 18, 2024.