Meeting Banner
Abstract #0074

Parameter-Free Compressed Sensing Reconstruction Using Statistical Non-Local Self-Similarity Filtering

Mariya Doneva1, Tim Nielsen1, Peter Brnert1

1Philips Research Europe, Hamburg, Germany

In this work, we present a CS reconstruction based on statistical non-local self-similarity filtering (STAINLeSS). The method provides improved image quality compared to wavelet based CS reconstruction and does not require any parameter adjustments. All the parameters are automatically determined by the noise estimation in the receive channels obtained from a standard noise measurement.

Keywords

achieves adaptive additional adjust aliasing alternating around artifact balance beginning best better body brain choice chosen clearly clinical coil combination combined comp computation computed constant constraint crucial degree density depend determined deviation difficult disk domain enables enforced estimation features fidelity filter filtering finite formula free head hies highly hods hold holding imaginary improved incoherent intro iterative iteratively lack lead learned local manually mated measure meters minimizing motivated necessarily need noise norm numb often onto optimizing particular patch patches peter phantom pixel pixels preserved pressed proposed quality real receive reconstruction reduced reduction regularization retrospectively self sense similarity space sparsity spatially stainless stand statistical statistically steps superior though thresh transform tuned variable varying visualization wavelet wavelets wing