A Real-Time Application FOR Removal of Non-Brain Tissues from MRI Images

Project Title

A Real-Time Application FOR Removal of Non-Brain Tissues from MRI Images

Project Area of Specialization Artificial IntelligenceProject Summary

Automatic removal of brain from magnetic resonance imaging (MRI) is a requirement in the field of Neuro image processing. The accuracy of various image processing applications lies on the effectiveness of skull stripping. Elimination of non-brain tissues can be an interesting exercise considering the numerous difficulties of human brain, unequal factors of MR scanners, modified unique feature and many more. Skull stripping is a beneficial technique for segmenting the brain tissue which is used for study of neuro imaging data. Thus, exact segmentation of brain tissue by removal of non-brain tissues like skull, muscle/skin, is an essential task for diagnosis a disease and pre-planning for a surgery. Digital image processing concepts is used in number of application and the scope of these application is increasing gradually in medical field for detection of brain anomalies. Brain MRI segmentation is an important task because it effects the result of the complete scanning. In past studies the skull remover technique is used for mapping of brain, segment brain tumor, and tissue classification. Skull removing is a major aspect in brain imaging applications and a process of removing non-brain tissue from a medical image of the brain (MRI scan). Several techniques have been proposed, manual or semi-automated methods are labor- intensive, operator-dependent, time consuming and thus are not desirable in large- scale studies. Image segmentation is an important step in many medical applications involving 3D visualization, computer-aided diagnosis, measurements, and registration. Therefore, the main objective of this project will develop an automated model for more accurately detection of the non-brain tissues with less false-positive rate.

Project Objectives

Objective of this project is to design a real-time application is to take an MRI image and remove the skull part and only target the brain tissues.and develop an automated model for more accurately detection of the non-brain tissues with less false-positive rate.

Project Implementation Method

• Matlab/Python (Programming Language)
• Edraw Max (Diagramming Software)
MS-Word (Creating Documents)
• MRI (Imaging-Technique)

Benefits of the Project

The trend of using modern technologies in the field of health is rapidly increasing.Manual detection of a brain disorder is time-consuming and error prone.To overcome this problem there is a need for advance monitoring technology to detect brain disorder.  

.Skull removing is one of the most challenging tasks on MRI images because the brain and nonbrain (skull) both images have same intensity which effect the accurately detection of the non-brain tissues with less false-positive rate.The objective of this project is to design a real-time application is to take an MRI image and remove the skull part and only target the brain tissues and give accurate result which is a grate contribution in a feild of medical sciences

Technical Details of Final Deliverable

the present study, a brain tumor detection approach is proposed based on Magnetic resonance imaging (MRI) running on Raspberry Pi (PRI) hardware. We’re going to pair our Raspberry Pi with the
Intel Movidius Neural Compute Stick coprocessor. The NCS Myriad processor will handle the more
demanding computations. Raspberry Pi is the first component selected to process the automatic vision
and deep learning standalone system. It is a 1GB credit-card sized and low-cost computer, which is
capable to process and send images wirelessly to MATLAB utilizing the PC’s GPU’s. It can also be used to
do onboard processing by using the Intel Movidius Neural Stick making it a gadget that can be deployed
easily. The process of offloading the most expensive deep learning task to the Movidius NCS frees up
the Raspberry Pi CPU to handle the other tasks. Each processor is then handling an appropriate load. We
are using more computationally efficient networks with a smaller memory/processing footprint such as
YOLOv2-SqueezeNet.

There are number of advantages and reasons; we used RPI in the project. Some are listed below 20
? Its compact size make it useful, portable and suitable for robotics
? Its video streaming capability
? Its operating system is Linux which in turn supports Open CV, Python programming and different
image processing techniques
? Its versatility in terms of functions and features.

Final Deliverable of the Project Hardware SystemCore Industry HealthOther IndustriesCore 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)	28500
raspberry pi 4	Equipment	1	28500	28500