Project Home
PDF access
Authors
Aly Yasser, Sandy Elias George, Anas El Shazly, Yasmina Basim,Dr.Ammar Mohamed, Eng. Nour El-Huda
Publishing Date
May 14, 2022
Abstract
Periodontal diseases are prevalent in both developed and developing nations, affecting between 20% and 50% of the global population. It’s the sixth most prevalent inflammatory illness in the world. The most common cause of tooth loss is periodontal disease. A more consistent technique of detecting Periodontal Disease is necessary owing to the limitations of the current diagnostic paradigm, human error that causes inaccuracy, and inconsistent judgments by various examiners. An automated medical support system is in high demand for screening at-risk individuals for periodontitis and early detection may postpone the beginning of tooth loss, particularly in small communities and health care settings with a shortage of dental specialists, due to advancements in machine learning research such as Transfer learning. Transfer Learning is a machine learning technique that allows data scientists to apply what they’ve learned from a previous machine learning model that was used for a similar task; as VGG16, VGG19,ResNet.
1.1 Purpose of this document
The purpose of this Software Requirements Specification document is to lay out the requirements for the software “A deep learning system[1] to diagnose and predict periodontal disease using X-rays.” The document will be useful to future users in the dental industry, such as dentists and dental students.
1.2 Scope of this document
This document illustrates the outcome of the software “A deep learning system to diagnose and predict periodontal disease using X-rays” will be a medical report that defines whether the provided image was periodontal or non-periodontal, and if periodontal, it will detect the severity level (Mild, Moderate, severe).
1.3 System Overview
At first the dentist enters a panoramic radio-graph image then it goes through the pre-processing of the image by resizing it then we enter the image segmentation phase after that we apply our deep learning method to check whether it’s periodontal or healthy. If appears that the patient is periodontal, we enter the decision maker part where we check for severity whether it is:
1. Mild
2. Moderate
3. Severe
1.4 System Scope
Our proposed system is designed to detect the periodontal disease and its severity(mild,moderate,severe). The system will detect the periodontal disease faster than normal. It aims to eliminate the use of dental probe as well as reduce the time and effort of the dentist. The system will:
• Detect if the radiograph image is healthy or unhealthy.
• Detect the periodontitis’ severity.
• Generate a report with the patients’ data and results