how to extract roi automatically

how to extract blobs (affected region) in lungs CT scan image without free hand selection using auto selection based on color r something . in that extracted output it needs to display only the blobs in white. background needs to be black

[Walter Roberson]

Is this cancer or TB to be detected, or something else?

CT scans are grayscale so you will not be able to detect anything by color.

[sathish kumar rb]

Cancer & Bronchities . I have attached my segmented lung

Answer 1

[Selva Karna]

use imroi

[Walter Roberson]

imroi is not able to automatically decide where the ROI should be.

Answer 2

[Image Analyst]

Look for papers describing how to do that here: http://www.visionbib.com/bibliography/contentsmedical.html#Medical%20Applications,%20CAT,%20MRI,%20Ultrasound,%20Heart%20Models,%20Brain%20Models or in PubMed.

[sathish kumar rb]

i this right r wrong

[Image Analyst]

Not sure what you want so I can't tell if it's right or wrong. I have a demo for segmenting the lungs - see attached and see if it works for you.

[sathish kumar rb]

i need to extract the roi(blobs) affected region only . the affected region displays like white and others will be black ...

[sathish kumar rb]

Input: Segmented Lung

Output: ROIs with class labels

The pathology bearing regions are the region of interests. ROI is used to identify the abnormality from the segmented lungs.

Algorithm: ROIs Extraction

1. Step 1: Extract the PBRs from the lung tissues as ROIs by finding the range of pixel intensity values to differentiate the PBRs and other lung pixels. If no PBRs present, then the slice is considered to be normal.
2. Step 2: Obtain the class label information of each ROI from an expert.

[Image Analyst]

What are PBRs?

Go here for better algorithms: Medical Image Processing Algorithms

Try imfreehand(), like in the attached demo, to have your expert draw the ROI.