WITHDRAWN: An advanced machine learning technique for the analysis of retina fundus images
| Autor: | Abhijit Bandyopadhyay, D.V. Ravi Shankar, Aatif Jamshed, Arun S. Tigadi |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
010302 applied physics
Artificial neural network Computer science business.industry Process (engineering) Deep learning 02 engineering and technology Fundus (eye) 021001 nanoscience & nanotechnology Machine learning computer.software_genre 01 natural sciences Visualization Teaching skills 0103 physical sciences Artificial intelligence 0210 nano-technology business computer |
| Zdroj: | Materials Today: Proceedings. |
| ISSN: | 2214-7853 |
| DOI: | 10.1016/j.matpr.2020.10.278 |
| Popis: | Each dataset with various artificial markers gives more or less good results. The type of alters that the network learns during the training steps tends to be very dependent. With Deep Learning, we will try to offer a response early to classify diabetic retinopathy and then attempt to visualise the important regions inside the classification fundus photos. This illustrates how a profound neural network distinguishes between various groups and whether this knowledge can aid early diagnosis or the detection of biomarkers. Firstly, the theoretical approach to the theory of deep learning and neural networks consists of a literature review and a discussion of the state of the art. The second is seen and used to test and research the visualisation process of Grad-CAM in the different retinal data sets with artificial markers. The findings and conclusions are eventually discussed. Deep teaching skills have been verified in this work, but more needs to be done to better understand how the knowledge of the qualified networks can be used, particularly for some medical images. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect