Reflection Ultrasound Computed Tomography (RUCT) is gaining prominence as an essential instrument for breast cancer screening. However, image quality is often compromised by the variability in sound speed across breast tissue. Traditionally, RUCT utilizes the Delay and Sum (DAS) algorithm, where the Time of Flight significantly influences image brightness, based on an oversimplified assumption of uniform sound speed. This study introduces a novel self-supervised deep learning model, the Self-Supervised Blind Segment Network, tailored to refine delay estimation in radio frequency (RF) data processing. Our approach tackles the issue of accurately estimating RF data delay by leveraging the adaptability of the model and enhancing accuracy through the spatial consistency in receiving arrays. Utilizing standard metrics like Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM), we evaluated our method’s effectiveness. The findings reveal that our approach yields substantial improvements, achieving an average PSNR of 20.04 and an average SSIM of 0.58, notably under conditions of sparse transmission. The conducted experimental analyses affirm the superior performance of our framework compared to alternative enhancement strategies. The code is available on github.com/HLUPUP/BSEGN.