Random math

Compressed sensing is a mathematical theory for extracting sparse signals from a sea of noise that has spawned inventions ranging from the single-pixel camera to facial recognition software. Now, a team from Rice University has applied the concept to infection diagnostics, creating an assay that can detect virtually any pathogen from a small set of random DNA probes. By not requiring probes to be tailored for each microbe, the assay offers the first practical solution to the need for a universally applicable infection diagnostic.

Compressed sensing mathematics was developed about a decade ago to efficiently reconstruct rare signals based on two essential properties: their “sparseness” relative to other signals and their “incoherence” with other signals, both of which allow them to be distinguished from noise and one another. ...