Ultrasound imaging offers numerous benefits compared to other modalities, including portability, affordability, and real-time volumetric imaging capabilities. However, its limited field of view hinders image interpretation, significantly limiting its potential clinical applications. This study focuses on creating tomographic ultrasound imaging by automatically registering partially overlapping 3-D ultrasound volumes (coarsely localized in space). This approach can be used to gather a comprehensive field of view of the anatomical area imaged and to enhance the acquired image information. Potential applications involve both tomographic/wide-view imaging using conventional ultrasound probes and future technologies, where multiple ultrasound probes will be able to scan simultaneously the anatomy of interest. This work describes the development of advanced artificial intelligence and image analysis methods to create real-time 3-D anatomical reconstructions of a designated area of interest. Simultaneously (or quasi simultaneously) acquired ultrasound volumes with partial overlap can be spatially aligned using rigid registration, because they present geometrically similar information. We used novel supervised and unsupervised transformers and compared the results with more traditional methods such as mutual information and cross-correlation. This application contributes to creating new imaging processing techniques for ultrasound imaging and making this complex modality more user-friendly and clinically usable.