Monitoring the growth of Polycystic Ovary Syndrome using Mono-modal Image Registration Technique
Suganya Ramamoorthy, R. E. Vinodhini, Rajaram Sivasubramaniam · 2019
In this paper, a unique approach to determine poly cystic ovary syndrome (PCOS) in female at her early stages caused due to an endocrine abnormality is proposed; ultrasound abdomen scan image is an efficient tool to determine PCOS. Along with it, the growth of the cyst is monitored at different intervals by image registration technique which affects females during their reproductive cycle. Initially, cyst detection is carried out by preprocessing technique and further the growth is monitored by image registration technique. To pre-process the scan image for speckle reduction, the existing approaches such as Gabor, Gaussian filter, adaptive filter, wavelet filter were used. These approaches have its own limitations and fail to detect subtle parts of speckles which contain the minute information of cysts. To overcome this, wavelet filter db2 is adopted in this paper. The growth of the cyst is being monitored at regular intervals by applying image registration technique with correlation coefficient similarity metrics and affine transformation. The system is implemented in Matlab. The proposed work will detect PCOS at the earlier stage with an accuracy of 93%. The advantage of this system is, to help the experts in the diagnosis of PCOS and deciding the necessary therapies regarding the patient's condition in its early stages.