Seismic data denoising has always been an indispensable step in the seismic exploration workflow. The quality of the results directly affects the results of subsequent inversion and migration imaging. In this article, we proposed a fast and flexible convolutional neural network (FFCNN) based on DnCNN. In contrast to the existing DnCNN and other artificial intelligence (AI)-based denoisers, FFCNN enjoys several desirable properties: 1) downsampling and upscaling operations, which can sensibly reduce runtimes and memory requirements while maintaining the denoising performance, and 2) we introduced noised level maps, which can make that a single convolutional neural network (CNN) model is expected to inherit the flexibility of handling noise models with different parameters, even spatially variant noises by noting $M$ can be nonuniform. Another benefit of increasing the noise-level map is that we can preserve more useful seismic data information by controlling the tradeoff of noise removal effect and seismic data detail preservation. For real seismic data denoised work, the main work and advantages of this article are concentrated on the following two aspects: 1) we introduced a data augmentation strategy to overcome the lack of well-labeled samples and 2) transfer learning has been introduced to the training processing, which used the well-trained synthetic seismic data denoising network as a pretrained model. In this way, we can greatly accelerate and optimize the learning efficiency of the training network. Ultimately, we can greatly improve the computational efficiency and denoising performance based on this intelligent denoised network FFCNN. Finally, numerical experiments prove the effectiveness of our method in synthetic and real seismic data

High-resolution images of human brain are critical for monitoring the neurological conditions in a portable and safe manner. Sound speed mapping of brain tissues provides unique information for such a purpose. In addition, it is particularly important for building digital human acoustic models, which form a reference for future ultrasound research. Conventional ultrasound modalities can hardly image the human brain at high spatial resolution inside the skull due to the strong impedance contrast between hard tissue and soft tissue. We carry out numerical experiments to demonstrate that the time-domain waveform inversion technique, originating from the geophysics community, is promising to deliver quantitative images of human brains within the skull at a sub-millimeter level by using ultra-sound signals. The successful implementation of such an approach to brain imaging requires the following items: signals of sub-megahertz frequencies transmitting across the inside of skull, an accurate numerical wave equation solver simulating the wave propagation, and well-designed inversion schemes to reconstruct the physical parameters of targeted model based on the optimization theory. Here we propose an innovative modality of multiscale deconvolutional waveform inversion that improves ultrasound imaging resolution, by evaluating the similarity between synthetic data and observed data through using limited length Wiener filter. We implement the proposed approach to iteratively update the parametric models of the human brain. The quantitative imaging method paves the way for building the accurate acoustic brain model to diagnose associated diseases, in a potentially more portable, more dynamic and safer way than magnetic resonance imaging and x-ray computed tomography.