SpiNet-QSM: Model-based Deep Learning with Schatten p-norm Regularization

for Improved Quantitative Susceptibility Mapping

This work proposed an effective model-based deep learning framework to solve the inverse problem of Quantitative susceptibility mapping (QSM) and it is a a Schatten p-norm-driven model-based deep learning framework for QSM with a learnable norm parameter p to adapt to the data. QSM provides an estimation of the magnetic susceptibility of tissues from magnetic resonance (MR) phase measurements. Estimation of the tissue magnetic susceptibility (source) from the measured magnetic field distribution/local tissue field (effect) inherent in the MR phase images was achieved by solving the inverse problem. [Detailed  Description]

GitHub: https://github.com/venkateshvaddadi/SpiNet-QSM

ISDU-QSMNet: Iterative Specific Denoising via Unshared Weights for QSM

Reconstruction

This work proposed an efficient model-based deep learning technique by leveraging the important parameters of the model-based deep learning for the QSM reconstruction. This work proposed a powerful model-based deep learning model for improving the QSM reconstruction with iterative specific denoiser via Unshared-weights.  [Detailed  Description]

GitHub: https://github.com/venkateshvaddadi/ISDU_QSMNet

Real-Time Median Nerve Segmentation in Ultrasonography

This work proposed an efficient, lightweight CNN model for the fully automated segmentation of the median nerve. It providing a throughput of 43 frames in a second. It supports the real-time median segmentation inference. This model can efficiently segment the median nerve from wrist to elbow along with respective CSA calculation in comparison with the manual tracing of nerve boundaries performed by expert sonographers. This proposed model is also clinically validated by testing on clinical data. An end-to-end clinical setup was made along with a Python based graphical user interface (GUI) interface for median nerve segmentation, which can give clinical assistance to the sonographers. This setup has already been made available in the Aster-CMI Hospital for clinical validation in real time.  [Detailed  Description]

GitHub: https://github.com/venkateshvaddadi/MNSeg-Net