Researchers have framed the design of undersampling masks for Cartesian magnetic resonance imaging (MRI) as a feature subset selection problem. By estimating mutual information and conditional mutual information from candidate masks and their training set evaluations, they developed a quadratic unconstrained binary optimization (QUBO) model. This model, solved using simulated and quantum annealing, determines a sampling mask by selecting a feature subset. The resulting reconstructions demonstrated lower mean root mean square error and higher mean structural similarity compared to conventional methods across various sampling rates. AI
IMPACT Introduces an information-theoretic approach to optimize MRI undersampling, potentially improving image reconstruction quality.
RANK_REASON The item describes a novel research approach using QUBO for MRI undersampling mask design, detailed in a paper. [lever_c_demoted from research: ic=1 ai=0.4]
Read on Hugging Face Daily Papers →
- Cartesian
- Conditional mutual information
- Fastmri
- Fourier
- magnetic resonance imaging
- mutual information
- quantum annealing
- qubo
- simulated annealing
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →