Hypothesis-Driven Classification of Materials Using Nuclear Magnetic Resonance Relaxometry

摘要

Technologies related to identification of a substance in an optimized manner are provided. A reference group of known materials is identified. Each known material has known values for several classification parameters. The classification parameters comprise at least one of T₁, T₂, T_1ρ, a relative nuclear susceptibility (RNS) of the substance, and an x-ray linear attenuation coefficient (LAC) of the substance. A measurement sequence is optimized based on at least one of a measurement cost of each of the classification parameters and an initial probability of each of the known materials in the reference group.