Breast cancer remains one of the most prevalent and life-threatening diseases among women worldwide. Accurate and early segmentation of tumors in ultrasound images plays a critical role in reducing mortality and guiding treatment. This paper presents a novel deep learning framework that combines an Attention U-Net with a Transformer architecture to enhance segmentation performance. The proposed fused model effectively captures both local spatial features and long-range dependencies, improving delineation of tumor boundaries. In contrast to conventional training paradigms, we propose a sequential learning strategy: the model is first trained on benign tumor images and subsequently fine-tuned on malignant cases. Experimental evaluations on the BUSI dataset demonstrate that our approach significantly outperforms traditional UNet, TransU-Net, and standalone Attention U-Net models. Notably, the fused model achieves superior performance in segmenting malignant tumors, validating the effectiveness of combining attention mechanisms with global contextual learning through the proposed incremental transfer learning approach.

Diagnosing fractures from medical images is challenging due to the limited availability of large, annotated datasets and the inherent variability across imaging modalities, such as CT and X-ray. Generating synthetic images that combine information from different modalities may enhance diagnostic accuracy. In this work, we propose a GAN-based multi-modal image fusion framework to generate synthetic X-ray images from CT scans. The generated images were evaluated both qualitatively and quantitatively by comparing them with real Xrays using metrics such as MSE, PSNR, and SSIM. To evaluate the effectiveness of the fused data, we trained a ResNet-18 classifier to differentiate between fractured and non-fractured knees, incrementally augmenting the original image with additional fused channels. The results showed a clear improvement in classification performance when fused modalities were included, particularly when two or three fusion outputs were combined. This approach demonstrates significant potential for advancing diagnostic tools in medical imaging, particularly when multimodal data is limited or unpaired.

In ultrasound imaging, sonographers are tasked with analyzing scans for diagnostic purposes; a challenging task, especially for novice sonographers. Deep Learning methods have shown great potential in their ability to infer semantics and key information from scans to assist with these tasks. However, deep learning methods require large training sets to accomplish tasks such as segmentation and object detection. Generating these large datasets is a significant challenge in the medical domain due to the high cost of acquisition and annotation. Therefore, data augmentation is used to increase the size of training datasets to create the needed variability for deep learning models to generalize. These augmentation methods try to mimic differences among scans that result from noise, tissue movement, acquisition settings, and others. In this paper, we analyze the effectiveness of general augmentation methods that perform color, rigid, and non-rigid geometric transformation, to empirically analyze and compare their ability to improve the performance of three segmentation architectures on three different ultrasound datasets. We observe that non-rigid geometric transformations produce the best performance improvement.

Based on non-ionizing radiation, ultrasound scanning is safe to image a specific region of the body repeatedly to identify and localize target anatomical structures during therapeutic and diagnostic procedures. However, it is labor intensive, and requires sonographers to have extensive experience to be able to identify and track these anatomical structures of interest, making the identification and tracking process highly prone to errors. In this paper, we propose a framework to autonomously detect, localize and track anatomical structures in ultrasound scans during scanning and therapeutic sessions in real-time. The proposed framework uses a segmentation-based convolutional neural network (CNN) to detect and localize the target anatomical structure within a scan. Concurrently, it uses an optical flow CNN to track the movement of this structure across frames to accurately guide therapeutic procedures. We tested the framework on detecting and tracking the Vagus nerve in ultrasound scans. It achieved state-of-the art localization and tracking accuracy with an average error of less than 1.25 mm for localization and 0.75 mm for tracking while maintaining an inference time of less than 35 ms.

In this letter, we propose a computationally efficient implementation of the phase shift plus interpolation (PSPI) seismic migration technique. The PSPI is among the well-known migration techniques in the frequency-wavenumber (ω - kx) domain that accounts for lateral velocity variations in the earth subsurface. However, the large number of complex multiplications required to perform PSPI makes this method computationally expensive. The proposed method replaces the use of complex multiplications in the frequency-wavenumber (ω - k√) domain with real additions, reducing the computational complexity greatly. It also replaces the real-by-complex multiplications with real multiplications. It does so by using the fact that the PSPI technique maintains the amplitude and shifts the phase of seismic wavefields in the passband region of the extrapolators while attenuating the amplitude and maintaining the phase of the wavefield in the evanescent region of the extrapolators. The proposed method was tested by performing 2-D poststack depth migration to the well-known SEG/EAGE Salt Model. The computational complexity in the ω - kxdomain was reduced by 76.9%, while producing an accurate migrated section of the salt model.

In this paper, we present a novel method for designing explicit wavefield extrapolators to perform both prestack and poststack depth migrations of seismic data. This method is achieved through the design of finite impulse response filters that perform the depth migration in the frequency-space (F-X) domain. The design method works on finding the regularized least square solution for the filter impulse response by using the Cauchy norm as the regularization cost function. This cost function offers an adaptive damping on the approximated filter impulse response coefficients yielding a more accurate approximation of the filter coefficients. It also insures the stability of the recursive migration process through which the designed filters will be used and prevent both over shooting and dampening of the wavenumber response of the filters designed. To test the designed filters, we then conducted poststack depth migration to the well-known SEG/EAGE salt model zero-offset data set, where subsalt structures were accurately migrated. The filters were also designed and used to perform prestack migration on the challenging Marmousi model data set successfully. The obtained results indicate that these filters can be used to perform stable and efficient prestack and poststack explicit wavefield extrapolations.