A SMOTE-Enhanced Hybrid VGG16–ResNet-50 Model for Automated Diabetic Retinopathy Detection

Muhammad Faris; Shahzaib  Khalid; Muhammad Dilawar Khan; Mir Farooq Ali; Muhammad Mansoor  Mughal; Tariq Javid

Authors

Muhammad Faris Hamdard University, Karachi, Pakistan https://orcid.org/0000-0003-0958-2265
Shahzaib Khalid Hamdard University, Karachi, Pakistan
Muhammad Dilawar Khan Hamdard University, Karachi, Pakistan
Mir Farooq Ali Marche Polytechnic University, Ancona, Italy
Muhammad Mansoor Mughal Hamdard University, Karachi, Pakistan/University of Houston, Texas, USA
Tariq Javid Hamdard University, Karachi, Pakistan

Keywords:

Diabetic Retinopathy, Deep Learning, Hybrid Model, Fundus Image Classification, Quantization-aware Training

Abstract

Diabetic Retinopathy (DR) occupies the top space among the preventable causes of vision loss across the globe with people who are diabetic presenting a higher number of victims. It suggests a hybrid deep learning network that consists of the VGG16 and ResNet-50 architecture to improve classification of the severity of DR based on the retinal fundus image. To fit the data model, a balanced and preprocessed dataset was used by applying data augmentation and Synthetic Minority Over-sampling Technique (SMOTE). Training was performed on input images which were normalized to 512 x 512 pixels and carried out over 25 epochs with a batch size of 32. The suggested model attained mean accuracy, precise, recall at 86%, 85%, 84% and F1- score at 85% respectively as compared to benchmark meaning that the model is capable of robust classification. Quantization-aware training was also used to maximize the computational efficiency of the model where the model now takes 95 milliseconds on average to process a single image, suitable to be deployed in a near real-time fashion (low resources). The hybrid model shows scalability with promise of inclusion in automated DR screening systems and will, therefore, provide an early solution to accurate diagnosis despite a few misclassifications that occurred due to the visual similarities between the DR stages.

DOI: https://doi.org/10.59564/amrj/03.03/013

Author Biographies

Muhammad Faris, Hamdard University, Karachi, Pakistan

Associate Professor, Department of Biomedical Engineering

Shahzaib Khalid, Hamdard University, Karachi, Pakistan

Student, Department of Biomedical Engineering

Muhammad Dilawar Khan, Hamdard University, Karachi, Pakistan

Service Engineer, Department of Biomedical Engineering

Mir Farooq Ali, Marche Polytechnic University, Ancona, Italy

PhD Scholar, Department of Information Engineering

Muhammad Mansoor Mughal , Hamdard University, Karachi, Pakistan/University of Houston, Texas, USA

Sr. Lecturer/PhD Scholar, Dept. of Electrical & Computer Engineering

Tariq Javid, Hamdard University, Karachi, Pakistan

Professor, Department of Biomedical Engineering

References

Bryan R, Kagadis C. Introduction to the science of medical imaging. Med Phys. 2011.

DOI: https://doi.org/10.1017/CBO9780511994685.008

Vlaardingerbroek T, Boer A. Magnetic resonance imaging: theory and practice. Springer Sci Bus Media. 2013.

DOI: https://doi.org/10.1007/978-3-662-05252-5

Mohammad H, et al. Brain tumor segmentation with deep neural networks. Med Image Anal. 2017;35.

DOI: https://doi.org/10.1016/j.media.2016.05.004

Haaf K. Informing patient surveillance for lung cancer survivors. J Thorac Oncol. 2022.

DOI: https://doi.org/10.1016/j.jtho.2021.12.003

Baghban R, et al. Tumor microenvironment and therapeutic implications. Cell Commun Signal. 2020.

DOI: https://doi.org/10.1186/s12964-020-0530-4

Arabahmadi M, et al. Deep learning for smart healthcare. Sensors. 2022.

DOI: https://doi.org/10.3390/s22051960

Faris M, Javid T, Mahmood K, Aziz D, Ali MF, Mughal MM. An enhanced U Net and CNN based tumor edge detection technique in MR images. Allied Med Res J. 2023.

DOI: https://doi.org/10.56530/amrj.v3i1.169

Faris M, Sohail F, Pepe C, Ali MF, Zanoli SM. Detection of Diabetic Retinopathy Using Deep Learning. Proc. 26th International Carpathian Control Conference (ICCC); 2025:1–4.

DOI: https://doi.org/10.1109/ICCC65605.2025.11022858

Neelum N, et al. Deep learning model for brain tumor diagnosis. IEEE Access. 2020.

DOI: https://doi.org/10.1109/ACCESS.2020.2978629

Lauriola I, et al. Deep learning in NLP. Neurocomputing. 2022.

DOI: https://doi.org/10.1016/j.neucom.2021.05.103

Mudasir G, et al. Ensemble deep learning: a review. Eng Appl Artif Intell. 2022.

DOI: https://doi.org/10.1016/j.engappai.2022.105151

Charnpreet K, Garg U. AI techniques for cancer detection. Mater Today Proc. 2023.

DOI: https://doi.org/10.1016/j.matpr.2021.04.241

Asra A, et al. Improved edge detection for brain tumor segmentation. Procedia Comput Sci. 2015.

DOI: https://doi.org/10.1016/j.procs.2015.08.057

Kim M, Lee BD. Effective boundary extraction in segmentation. IEEE Access. 2021.

DOI: https://doi.org/10.1109/ACCESS.2021.3099936

Mawaddah H, et al. DL for brain tumor detection. In: ICOSNIKOM, IEEE. 2022.

DOI: https://doi.org/10.1109/icosnikom56551.2022.10034876

Neha B, et al. CNN for brain tumor classification. In: IMPACT. 2022.

DOI: https://doi.org/10.1109/IMPACT55510.2022.10029043

Muhammad A, et al. DL for brain tumor classification. Comput Electr Eng. 2022.

DOI: https://doi.org/10.1016/j.compeleceng.2022.108105

Ayesha Y, et al. Brain tumor analysis using VGG-16 ensemble. Appl Sci. 2022.

DOI: https://doi.org/10.3390/app12147282

Rao P, et al. Novel DL method for brain tumour detection. Biomed Signal Process Control. 2023.

DOI: https://doi.org/10.1016/j.bspc.2022.104549

Gulshan V, et al. Development and validation of a deep learning algorithm for detection of DR. JAMA. 2016.

Kaggle. EyePACS Dataset. Available from: https://www.kaggle.com/c/diabetic-retinopathy-detection/data

Decenciere E, et al. Feedback on a publicly distributed image database. Health Inf J. 2014.

Aburass S, Dorgham O, Al Shaqsi J, Abu Rumman M, Al-Kadi O. Vision Transformers in Medical Imaging: A Comprehensive Review of Advancements and Applications Across Multiple Diseases. J Imaging Inform Med. 2025; Mar 31.

DOI: https://doi.org/10.1007/s10278-025-01481-y

Huo G. Deep Learning Models for Diabetic Retinopathy Detection: A Review of CNN and Transformer-Based Approaches. In: Proceedings of the 2nd International Conference on Data Analysis and Machine Learning (DAML), Vol 1; 2024. p 594–598.

DOI: https://doi.org/10.5220/0013533700004619

Kaggle. Diabetic Retinopathy Datasets. Available from: https://www.kaggle.com/datasets/sovitrath/diabetic-retinopathy-2015-data-colored-resized.