Ultrasound tomography technology is rapidly emerging as a focus of medical imaging due to its advantages of non-invasiveness, cost-effectiveness, lack of radiation harm, and portability. However, when ultrasound waves propagate in media with high acoustic impedance contrast such as the musculoskeletal system, complex scattering often occurs, leading to waveform distortion and amplitude attenuation of the received signals, thus limiting the resolution and accuracy of reconstruction. Therefore, we exploit a circular to collect full matrix data for musculoskeletal tissues ( including numerical and in vivo examples), enabling collection of ultrasound signals in various modes including reflection, transmission, and multiple scatterings under full-aperture conditions. Subsequently, multiple algorithms are employed to reconstruct the qualitative and quantitative images of the target. Delay-and-sum technique is utilized to achieve structural imaging of strong reflection interfaces, while time-of-flight tomography imaging aids in reconstructing macro sound speed distribution images. Furthermore, full waveform inversion generates higher-resolution images of sound speed distribution by iteratively optimizing on top of time-of-flight tomography imaging results. Through numerical and experimental tests, the method combining time-of-flight tomography and multi-scale frequency-domain full waveform inversion is validated to precisely reconstruct different tissue components, such as skin, fat, muscle, and bones, achieving 0. 4 mm resolution. The research expands the application scope of ultrasound in medical imaging holding significant clinical value for accurate diagnosis of musculoskeletal disorders.