A method for the extraction of an object from a noisy background

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Image Processing

Background:

• Radioisotope imaging suffers from low signal-to-noise ratios due to scattered photons and background radiation.
• Image noise, resulting from low photon detection, complicates accurate object edge delineation.
• Abnormalities in medical images can present with low pixel values, mimicking background noise and further challenging segmentation.

Purpose of the Study:

• To develop an advanced image processing technique for precise object edge detection in noisy medical images.
• To enhance the accuracy of computer-aided diagnosis systems by improving image segmentation.
• To address the challenges posed by low signal-to-noise ratios and complex noise patterns in radioisotope imaging.

Main Methods:

• Development of a novel image processing technique utilizing 'moving window operations'.
• Integration of the technique within a computer-aided diagnosis research project involving artificial neural networks.
• Testing the developed method on male lung images acquired using a gamma camera.

Main Results:

• The image processing technique demonstrated effectiveness in determining object edges within noisy images.
• The method shows promise in distinguishing objects of interest from background noise, even in challenging cases.
• Successful application of the technique to gamma camera lung images indicates its potential clinical relevance.

Conclusions:

• The developed image processing technique offers a robust solution for edge detection in low signal-to-noise medical images.
• This advancement is vital for improving the performance of computer-aided diagnosis systems.
• The technique's successful validation on lung images highlights its applicability in nuclear medicine and diagnostic imaging.