Glaucoma Detection using Deep Learning.pptx
- 2. Presented By Fiza Fatima Usama Aziz Zunaira Asif 20-UON-0574 20-UON-0598 20-UON-0564
- 3. Introduction ● Computer vision has became the most widely used field of Artificial intelligence. It has made its way into the medical field too. ● Nowadays artificial intelligence is used in medical field to diagnose a disease, optimize the treatments and many more. ● Deep learning algorithms are being used to make diagnostic models to detect the disease at an early stage.
- 4. Background ● Glaucoma is an eye disease which is caused when the nerve between brain and the eye connected is damaged. Glaucoma must be detected at an early stage because it can cause permanent blindness if not treated early ● Glaucoma mostly occurs at people aged over 40. It is the most common cause of blindness in the elderly people. Over 80 Million people are living with glaucoma world wide by the year of 2024 this number is expected to go upto100 Million in the next 20 years.
- 5. Problem Statement ● Glaucoma is the most commonly found eye disease and it is the most common cause of blindness in elderly people ● It must be detected early. To cure the disease it is most important to find and start its treatment as early as possible. ● Sometimes it is very easy for an Ophthalmologists to oversee glaucoma in the eye. A trained eye can also make mistakes while detecting glaucoma and they can overlook it ● This is where we need to take help from technology and solve this problem by using technology on it.
- 6. Proposed Methodology ● We proposed a solution which is to make a deep learning model to help the Ophthalmologists to an extent so that they can easily detect glaucoma in the eye. In our training phase we will train the model using a proper dataset. ● CNN algorithm is most common deep learning algorithm in which neural layers are formed by which features are extracted from the images or different data provided ● CNN is a totally unbiased algorithm so to add biasness in the algorithm to check a specific area of the image data we use different spatial attention algorithms which will help us increase the accuracy of the model.
- 7. Dataset The dataset we are going to use for the learning phase contains 100,000+ color fundus images from 60,000+ patients. All images were assigned by human experts with the labels referable glaucoma, no referable glaucoma, or ungradable.
- 8. Diagram
- 9. Scope Glaucoma is the most common disease found in the elderly people mostly over 40. According to the WHO, glaucoma affects 4.5 million people globally. As people get older, glaucoma becomes more common.
- 10. Previous Work We have read different research papers and get to know about the previous work done in the field What methodology they used what results they got using the methodology they proposed.
- 11. Literature Review The researcher introduced a model to discover the glaucoma following the performance of preprocessing the researcher separated the optic cup and optic disk by pixel-based threshold and finally he computed the CDR by distributing the cup pixel by the disk pixels this framework doesn’t perform well if the sample is changed
- 12. Literature Review A research in which the researcher introduced a Machine Learning based model for glaucoma detection. To segment the blood vessels, the researcher used the canny edge detection method. In second step the Finite Element Modeling was used to extract the features from the images to divide the healthy eye and glaucoma eye. The work performs well on the noisy data but requires evaluation on complicated dataset.
- 13. Literature Review The researcher developed a technique that pulls information from three models: AlexNet, ResNet-50, and ResNet-152, which were all combined to predict glaucoma. This technique does better glaucoma classification, but the framework is computationally expensive.
- 14. Tools ● Visual Studio ● Python (Language) ● Figma (Design And Models)
- 15. Conclusion In Conclusion we will be making a Deep learning model which will help in detecting the glaucoma in the eye as it is necessarily to detect the glaucoma at its early stage to cure it completely.
- 16. Thank You!