Multiscale superpixel method for segmentation of breast ultrasound

Abstract Background In medical diagnostics, breast ultrasound is an inexpensive and flexible imaging modality. The segmentation of breast ultrasounds to identify tumour regions is a challenging and complex task. The major problems of effective tumour identification are speckle noise, artefacts and low contrast. The gold standard for segmentation is manual processing; however, manual segmentation is a cumbersome task. To address this problem, the automatic multiscale superpixel method for the segmentation of breast ultrasounds is proposed. Methods The original breast ultrasound image was transformed into multiscaled images, and then, the multiscaled images were preprocessed. Next, a boundary efficient superpixel decomposition of the multiscaled images was created. Finally, the tumour region was generated with the boundary efficient graph cut segmentation method. The proposed method was evaluated with 120 images from the Thammassat hospital database. The database consists of 30 malignant, 30 benign, 60 fibroadenoma, and 60 cyst images. Popular metrics, such as the accuracy, sensitivity, specificity, Dice index, Jaccard index and Hausdorff distance, were used for the evaluation. Results After careful experiments, the results indicate that the proposed method achieves high segmentation accuracy with 97.3% for other benign, 94.2% for malignant, 96.4% for cyst and 96.7% for fibroadenoma breast ultrasounds. The results validate that the proposed model can achieve higher segmentation accuracy on the dataset than state-of-the-art segmentation methods. Conclusions This paper presents a method that segments the tumours in breast ultrasounds, and the results indicate that the proposed method performed better than several state-of-the-art methods. The algorithm used will act as a tool for the effective segmentation of breast ultrasound images.