Development and testing of digital image processing techniques in diffuse optical tomography system

Objective: Continuous Wave Diffuse Optical Tomography (CWDOT) system is one of the imaging modality used in the medical. The aim of this study is to apply different three dimensional (3D) image processing methods to breast phantom images and determine the most appropriate image processing methods that give the most realistic image in CWDOT. Methods: In the study, the measurements were acquired on breast phantoms in CWDOT. The first phantom was made with a mixture of intralipid, water ,and Indocyanine green (ICG), an inclusion (tumor-like structure) to represent tumor was placed to represent the tumor. The second phantom was prepared by placing the spleens (tumor-like structure) into the tail fat to obtain a more realistic breast phantom. First, Image segmentation method which is one of the global and local thresholding method was applied in the image processing method. Then spatial filter; Average, Gaussian, Laplacian, Laplacian of Gaussian (LoG) filtering methods were implemented. Final, images were obtained by applying nearest, bilinear, bicubic and cubic spline interpolation methods. Results: The results of image processing were compared with visual and quantified evaluation. The performance of image processing algorithms was compared using Peak signal-to-noise ratio (PSNR), Mean square error (MSE) ,and Structural similarity index (SSIM) methods. The most appropriate image processing algorithms weere defined base on positions of tumor-like structures within the breast phantoms were displayed true shapes and sizes. Conclusion: Images were obtained clearly ,and accurately using the Otsu method, Gaussian filtering ,and bi-cubic interpolation methods in CWDOT system. This thesis will contribute to researchers in using the most convenient image processing algorithms who will perform in-vitro and clinical experiments in this field. Key words: diffuse optical tomography, image processing, image segmentation, image filtering, image interpolation.