Identification of Mild Cognitive Impairment Dementia

Project Title

Identification of Mild Cognitive Impairment Dementia

Project Area of Specialization Artificial IntelligenceProject Summary

The purpose of this project to identify Mild Cognitive Impairment (MCI) an early stage  of Alzheimer’s Disease (AD).  The growth rate of AD has increased than cancer round the globe.  It affects human’s cognitive ability and memory.  It has no cure, once a person is affected. However, if early symptoms are diagnosed timely then it can be avoided or its growth rate could be slow down using medication. An early symptom indication of AD is called Mild Cognitive Impairment (MCI).  Lack of distinct medical cure for this disease has urged the need to identify factors of early diagnosis of AD using the automated process. Its common symptoms include cognitive impairment, memory loss, difficulty in thinking and understanding words, and inability to identify family members. Moreover, the lack of Beta proteins and CDR [(Clinical Dementia Rating) ]code in MRI (Magnetic Resonance Imaging) scans are important Clinical data parameters. Identification of these parameters using MRI scans of ADNI dataset is less prone to errors as MRI generates quality medical images. This project aims to diagnose MCI using CNN (Convolution Neural Network) applied to MRI scans.

Project Objectives

• The objective of this identification system is to predict the earliest signs of MCI and to track the disease using biomarkers like Structural MRI.
•  Train the Algorithm that provides better accuracy results from the dataset of images data of different centers in order to reach both reliability and reproducibility of results.

Project Implementation Method

In this project our approach towards the project would be "Agile" as agile is the software development under which requirements and solutions evolve through the collaborative effort of self-organizing and cross-functional teams. In this we will be going through the following steps:

1. Data acquisition
2. Preprocessing of dataset
3. Training the dataset
4. K-fold cross-validation of the dataset
5. Test cases
6. Model performance evaluation
7. Web services development
8. Integration of trained mode with web services application
9. Final integration and testing

Benefits of the Project

This will help doctors and patients to predict the early sign of MCI and to manage their health regime. As it is a neuro-generative disease and at present, there is no cure for Alzheimer's, the only cure is prevention. so this application will be helpful to early diagnose the symptoms of the disease at a mild stage, i.e., MCI; so that medication could be effective. 

Technical Details of Final Deliverable

Neural Network:

We will train our machine learning model using a convolution neural network. 

Data set :

The Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset is available for the research community. This dataset includes MRI scan images for Alzheimer's, cognitive impairment subjects, and elderly controls.

Tools and Technology :

HTML, CSS, Bootstrap, web services for web application

SQL, FLASK Framework, Python 3, Pycharm notebook, Sickit learn Machine learning Library, Numpy, PANDAS, matplotlib, openCV etc.

Final Deliverable of the Project Software SystemCore Industry MedicalOther Industries Health Core Technology Artificial Intelligence(AI)Other TechnologiesSustainable 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)	10000
domain hosting	Miscellaneous	1	5000	5000
Azure services	Miscellaneous	1	1000	1000
machine learning course	Miscellaneous	2	2000	4000