Classification of Diabetic Retinopathy and Normal Retinal Images using CNN and SVM

Zaman²,

et al. 2021

Electronics

Internet of Things (IoT) communication technologies have brought immense revolutions in various domains, especially in health monitoring systems. Machine learning techniques coupled with advanced artificial intelligence techniques detect patterns associated with diseases and health conditions. Presently, the scientific community is focused on enhancing IoT-enabled applications by integrating blockchain technology with machine learning models to benefit medical report management, drug traceability, tracking infectious diseases, etc. To date, contemporary state-of-the-art techniques have presented various efforts on the adaptability of blockchain and machine learning in IoT applications; however, there exist various essential aspects that must also be incorporated to achieve more robust performance. This study presents a comprehensive survey of emerging IoT technologies, machine learning, and blockchain for healthcare applications. The reviewed articles comprise a plethora of research articles published in the web of science. The analysis is focused on research articles related to keywords such as `machine learning’, blockchain, `Internet of Things or IoT’, and keywords conjoined with `healthcare’ and `health application’ in six famous publisher databases, namely IEEEXplore, Nature, ScienceDirect, MDPI, SpringerLink, and Google Scholar. We selected and reviewed 263 articles in total. The topical survey of the contemporary IoT-based models is presented in healthcare domains in three steps. Firstly, a detailed analysis of healthcare applications of IoT, blockchain, and machine learning demonstrates the importance of the discussed fields. Secondly, the adaptation mechanism of machine learning and blockchain in IoT for healthcare applications are discussed to delineate the scope of the mentioned techniques in IoT domains. Finally, the challenges and issues of healthcare applications based on machine learning, blockchain, and IoT are discussed. The presented future directions in this domain can significantly help the scholarly community determine research gaps to address.

Section: Title Accuracy Relative Demeritsmentioning

confidence: 99%

Comprehensive Survey of IoT, Machine Learning, and Blockchain for Health Care Applications: A Topical Assessment for Pandemic Preparedness, Challenges, and Solutions

Zaman²,

et al. 2021

Electronics

“…Qomariah et al used Support Vector Machine (SVM), using the CNN as the input features for classification. They got the highest accuracy values of 95.83% and 95.24% for base 12 and base 13 [6]. Mobeen-ur-Rehman et al used 3 different pre-trained CNN models: AlexNet, VGG-16, and SqueezNet.…”

Section: Related Workmentioning

confidence: 99%

Finding the Best Performing Pre-Trained CNN Model for Image Classification: Using a Class Activation Map to Spot Abnormal Parts in Diabetic Retinopathy Image

Jihyung¹

2021

AJBLS

Diabetic retinopathy (DR) is a common eye disease that people get from diabetes. About 33.7% of the people with diabetes have DR. With our datas, which are pictures of the eyeball with and without DR, we tried different convolutional neural network (CNN) models to get the best accuracy score. We tested our datas with a default CNN model, and 5 different pre-trained models: MobileNet, VGG16, VGG19, Inception V3, and Inception ResNet V2. The default CNN model didn't perform very well, getting only 10.4%. The pre-trained model also didn't perform as good as expected, so we decided to use GRU with the models, which increases the score. For the higher accuracy, we added bidirectional GRU to train the whole parameters in the model. The 5 different pre-trained models scored an average of 74.2% accuracy score, and Inception ResNet V2 with bidirectional GRU included scored the highest accuracy, achieving 83.57%. For additional study, we used a class activation map to spot the abnormal parts of the eyeball with DR, and we could spot abnormal veins and bleeding on the eyeball. However, our research has limitations on that we did not use the segmentation methods, which is more advanced technique compared to classification, such as U-net, Fully Convolutional Network (FCN), Deep Lab V3, and Feature Pyramid Network. Furthermore, even though our model classified 5 different classes, the fact that the highest accuracy score was lower than 90% is also a limitation. For further study, we would prepare a masking method for applying segmentation methods to our dataset.

“…Pada penelitian ini hanya dilakukan klasifikasi 2 kelas antara normal dan severe NPDR yang tentu saja akan sangat mudah untuk ditemukan cirinya, sehingga nilai akurasi akan sangat tinggi. (Qomariah, 2019).…”

Section: Pendahuluanunclassified

Deep Learning untuk Klasifikasi Diabetic Retinopathy menggunakan Model EfficientNet

et al. 2020

ABSTRAKDiabetic Retinopathy merupakan penyakit yang dapat mengakibatkan kebutaan mata yang disebabkan oleh adanya komplikasi penyakit diabetes melitus. Oleh karena itu mendeteksi secara dini sangat diperlukan untuk mencegah bertambah parahnya penyakit tersebut. Penelitian ini merancang sebuah sistem yang dapat mendeteksi Diabetic Retinopathy berbasis Deep Learning dengan menggunakan Convolutional Neural Network (CNN). EfficientNet model digunakan untuk melatih dataset yang telah di pre-prosesing sebelumnya. Hasil dari penelitian tersebut didapatkan akurasi sebesar 79.8% yang dapat mengklasifikasi 5 level penyakit Diabetic Retinopathy.Kata kunci: Diabetic Retinopathy, Deep Learning, CNN, EfficientNet, Diabetic Classification ABSTRACTDiabetic Retinopathy is a diseases which can cause blindness in the eyes because of the complications of diabetes mellitus. Therefore, an early detection for this diseases is very important to prevent the diseases become severe. This research builds the system which can detect the Diabetic Retinopathy based on Deep Learning by using Convolutional Neural Network (CNN). EfficientNet model is used to trained the dataset which have been pre-prossed. The result shows that the system can clasiffy the 5 level of Diabetic Retinopathy with accuracy 79.8%. Keywords: Diabetic Retinopathy, Deep Learning, CNN, EfficientNet, Diabetic Classification