A non invasive system for pre diagnosing of human digestive disorders

Project Title

A non invasive system for pre diagnosing of human digestive disorders

Project Area of Specialization Biomedical EngineeringProject Summary

It is a non-invasive system for diagnosing human digestive system disorders by taking human iris image is an input to find out the disorders of the human digestive system.

According to the study of iridologist, when any change occurs inside the body other than regular activities the mark or point appear in the iris.
That changed in the iris such as mark or point is the description of the particular body organ or system. These changes in iris help to identify the specific disease or disorder in a system through the Iridology map and machine learning algorithm.

so, first of all, we will train our system through a convolutional neural network (CNN)  by giving the healthy subjects iris pic and digestive disorder subject iris pic.

The system works in a real-time environment and processes the algorithm on a CPU-GPU based machine.

Project Objectives

As we know the conventional tools and procedures for diagnosing of digestive disorders are expansive and time-consuming, and also we know that nowadays due to the bad nutrition every third person having a digestive problem, therefore we need a fast and cheap accurate diagnosing system which is beneficial for the patients to save his precious time & money.

conventional procedures like (Endoscopic procedures, image tests, lab tests).

Project Implementation Method

For input, we have an Iridology Camera with an image resolution of  2560x1920 pixels.

For processing system, we have Hardware Architecture which holds multiple processing nodes.
In the current project, we practiced three nodes based high-performance computing system.
Each node of the system Architecture utilizes the multi-RISC processor and GPU-accelerated cores, which comprises General Purpose (Intel/AMD) multi-core processor and Graphical Processing Unit GPU.
Each node uses an Intel Xeon X5550 general-purpose processor and Nvidia GTX1080 GPU having 2560 Cuda cores. GPUs are used to reduce the cost and power conception and give data-level parallelism.
The IPSDD system architecture uses a Linux based operating system and uses open source and easy to program artificial intelligent frameworks such as (Tensorflow and CAFFE etc.).

And in the end, we will have an LCD screen to show us the results.

Benefits of the Project

This system provides us a lot of benefits than conventional tools and procedures.

This system cost cheaper than other conventional tools, so everyone can afford it.

There are no side effects of using this system.

It works faster to save patients precious time.

This system is very easy to use, so no expertise needed to use this system.

Technical Details of Final Deliverable

we already had brought the system in one peace just need to collect the iris pics of healthy and unhealthy subjects, and then we have to train our system through a Convolutional neural network (CNN). 

InshaAllah it will be ready sooner than expected date. At the end of March, our project will be in a complete running form.

Final Deliverable of the Project Hardware SystemCore Industry HealthOther Industries Medical 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)	74500
Intel Xeon X5550 general-purpose processor	Equipment	1	20000	20000
Nvidia GTX1080 GPU having 2560 Cuda cores	Equipment	1	42000	42000
misc (all included)	Miscellaneous	1	5000	5000
5 inch LCD HDMI Touch Screen Display TFT LCD	Equipment	1	4000	4000
19 inces LCD	Equipment	1	3500	3500