Digital Health Avatar

Project Title

Digital Health Avatar

Project Area of Specialization Artificial IntelligenceProject Summary

The Digital Health Avatar application has developed intending to detect human disease and shows human health to them so they can easily understand in which condition they are, and the gif avatars show by their actions how their health is right now. We build an android application. This application examines four types of disease including real-time facial stroke, picking an image from the gallery, heart rate, and blood pressure and their results alert the person with gif Avatars whether he is fine or not. If the result is not healthy, then the person can share this result with the doctor. We have focused on using avatars for those who are older and do not understand the measurements of the results that’s why we are using avatars that they understand easily. And we also kept in mind that social distancing is very important for a pandemic so people can check the disease through the application.

The Digital Health Avatar  check the human condition on blood pressure, heart rate and palsy face a stroke symptom and that will help them to care about their conditions before health will become worse. Mostly Humans cannot understand the medical measurements or terms so we can help them to understand by showing them avatars  help them to understand their conditions. If the result is not healthy what will happen if do not consider your health condition and by app, data user can send health record to the doctor about the health condition. The purpose of Digital Health Avatar checks human health and show the result on the spot and using the resulting app will alert the human that user condition according to result is healthy or not.

Project Objectives

The objectives of the thesis are shown as following:

i) To facilitate users to check the disease.

ii) To facilitate users to share medical records.

iii) To facilitate users to perform the risk assessment.

iv) To facilitate users to understand health conditions with the help of avatars.

v) To facilitate users who cannot afford doctors’ fees or test

fees

Main objective is to make a application that can be possible to detect disease maybe it difficult by mobile phone but it can  be possible by detecting the disease based on symptoms.

Project Implementation Method

Digital Health Avatar is a Mobile based application therefore front-end, and back-end will be developed in Android Studio. It will be developed using feature-driven development in which first we will build an overall model then build a feature list and then we will plan, design, and develop by feature. The project is only implemented to make a mobile app so it might take time to open because of synchronizing database. Since the application fetches data from the database SQLite and that no data connection is required. It is an android mobile application; the user of the application must have an android smartphone to access this application.

Performance Requirements  

Android mobile application  developed as a less APK mobile application so that it can work on almost any Mobile even with lower APK mobile.

Security Requirements The system should provide basic security features like username and password authentication for the database. 

Software Quality Attributes

Reliability

igital Health Avatar is a reliable mobile application. The application will be available in every case while the network goes down or not system will be available all time. Any personal information like phone numbers and addresses will be kept safe.

Maintainability The developer will maintain the mobile application data and all the services

Availability The mobile application will be all time available. Reusability Provided mobile application has reusability functions. Like when a user or other actors’ sign-ups, he/she can use the same profile for orders in the future.The model again can train with different disease images and can integrate in mobile phone.

Benefits of the Project

Digital health avatar is an application that allows to user maintain good health by examining the diseases whenever they want without wasting time and without paying any charges

This application help to find out the paralyzed face, blood pressure, and heartbeat rate.

Users can be able to check the paralyzed face in real-time in which the camera is on and give real-time results and pick the image from gallery in which select an image from gallery then classify which gives the result in percentage and also can check the blood pressure and heart rate by place the finger on the back camera which take the waves from the finger and also use the prediction values like age, weight, and gender in which place finger on the back of the mobile camera while flashing is turned on which predict the disease rate and show action thorough animated GIF avatars which represent the user-health is fine or not.

 There is no android-based application for the user to find the paralyzed face so with the help of Digital Health Avatar user can detect the facial paralysis that occurs during a stroke, and you don’t need the internet to run this application.

Doctors or anyone can use it for their use and there is not only these scope we want to make it further advance for the different diseases like Diabetes Retinopathy ,Acne , Rosacea disease detection based on the symptoms on external body.

Technical Details of Final Deliverable Model Experiments

Figure 1 : Normal Face Graph

Fig 1 shows the Normal face graph from dataset that is used for the training model that is used in the application

 

Figure 2 :Palsy face Desktop Result

Fig 2 shows the comparison between Normal and Palsy Face on desktop Realtime detection from the application and camera must working properly.

Figure 3: Normal face DesktopResult

Fig 3 shows the comparison between Normal and Palsy Face on desktop Realtime detection from the application and camera must working properly.

Figure 4: Homepage

Fig 4 after sign in user go to the homepage and where the user sees main functionalities buttons of the application.

          

Figure 5 Detection Page

Fig 5 shows the Detection screen of the application when user click the heart rate or blood pressure disease check then see this type of screen which give the instructions.

Figure6 : Measuring Page

Fig 6 shows the finger signal wave when user detect the heart rate and blood pressure and place finger on the back camera then show this type of screen. 

Figure 7 : Result Page

Fig 7shows the Result screen when user select the heart rate disease and check the disease then show this type of screen according to result of the application.

Figure 8 : Select Image from Gallery

Fig 8 shows the screen when click on the homage from check the palsy image from gallery then show this type of page where the user can upload the images.

Figure9 : Gallery Image Normal Face Result 1

Fig 9 shows the picture is picked from gallery by clicked the choose image button then the image was appeared and by clicked the button classify image the result shows 99.9 % and status 1 represent Normal face. 

Figure 10: Palsy face Realtime Result

Fig 10 shows the Realtime result of the palsy faces when user click to check the real time palsy face and real-time detection show this face is palsy with 99.17 percentage.

Data Gathering

Because of covid and rare cases of the stroke disease is about 250 to 271 in 100000 in Pakistan in last studies that conduct in Khyber Pakhtunkhwa (KP) and out of 22,500, 11,556 (51.4 %) and 10,944 (48.6 %) were females are so, we have created a dataset by taking pictures of normal and palsy faces from internet. The figure shows the folders of normal and palsy faces.

  

Figure 12 : Data Gathering of Normal Faces  Fig 12 shows the dataset consist of normal faces Images that is usedfor the model as training dataset and helps in detection of Palsy faces

 

Figure13 : Data Gathering of Palsy Faces

Fig 13 Images shows the dataset consist of normal faces Images that is usedfor the model as training dataset and helps in detection of Palsy faces.

Camera Sensors Use the camera sensors which is built-in android phones to detect the Blood pressure and heart rate place the index finger on the back camera to check the status of blood pressure and heart rate. Avatars  We have used gif animated Avatars which will indicate the state of the disease whether the user is fine or not and some gif avatars actions show how to use the app. Final Deliverable of the Project Software SystemCore Industry HealthOther Industries IT Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Good Health and Well-Being for PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	70000
Subscriptions	Equipment	1	70000	70000