MEDICAL IMAGE SEGMENTATION USING KENNEL PRINCIPAL COMPONENT ANALYSIS
EXPLANATION: The project should have the following segment
• Image Registration
• Image Segmentation
The visual representation of the project we are working at, for example when an Xray of a person/patient is been taken and the film is taken to the Doctor, what the Doctor does is that He takes a visual look at the film in order to see if there is any contour. It simply implies that the doctor has the correct form of the particular image He is viewing on the film in his brain. This will enable the doctor to detect where the error is. In Specialized word we will say the Doctor’s brain is the Database, so that whenever an Xray of a Patient is brought to him, the Doctor then compare the correct form of the Image in His Brain (Database) with the Improper one on the film (Source Image) to know whether the Source image has a Fault or not. What we are working at, is that we want to introduce Information and Computer Technology into it. Instead of the Doctor taking a visual look the Computer should stand in the gap.
THE REQUIRED COMPONENTS
KENNEL: ACT AS THE DATABASE (Where the correct form of the Medical Images will be stored)
We will be working with MRI Images and CT Images. In the Medical Field every image has an MRI (Magnetic Resonance Image) view and the CT (Computed Tomography) Image view. It simply means that for example if we take the Image of the Brain, it can be taken from an MRI source and a CT source. The Combination of the MRI and CT of the Image of the Brain gives the normal brain image of the particular person.
FUNCTIONS OF THE REQUIRED COMPONENTS
IMAGE REGISTRATION: It will register the MRI and the CT image together.
KENNEL: For example, if we are working on the Image of the Brain, In the Kennel there must be a correct form of the image of the brain. Similar to the illustration above that the Doctor’s brain is the Kennel and it contains the correct form of medical images
IMAGE SEGMENTATION: After Registration has taken place, there will be a comparison of the registered image and the correct form of that type of image in the kennel. The region where there is error is been detect and the particular region is been segmented.
 WE NEED TO DO IMAGE REGISTRATION AFTER THAT IMAGE SEGMENTATION.
 WE NEED TO KNOW THE TIME AFTER THE SEGMENTED IMAGE HAS BEEN COMPARED WITH THE IMAGE IN THE KENNEL
 WE NEED TO KNOW WHETHER THE IMAGE IS CORRECT
 WHAT IS THE TRESHOLD VALUE THAT IT USED TO RECOGNISE THE IMAGE IN THE KENNEL.
 TIME OF SEGMENTATION.
 THE IMAGE SEGMENTATION TECHNIQUE MUST BE REGION BASED USING GEOMETRIC ACTIVE CONTOUR AND NOT EDGED BASED.