Sampling and samples Chemical imaging

1 sampling , samples

1.1 detection limit
1.2 data analysis

1.2.1 software

sampling , samples

the value of imaging lies in ability resolve spatial heterogeneities in solid-state or gel/gel-like samples. imaging liquid or suspension has limited use constant sample motion serves average spatial information, unless ultra-fast recording techniques employed in fluorescence correlation microspectroscopy or flim observations single molecule may monitored @ extremely high (photon) detection speed. high-throughput experiments (such imaging multi-well plates) of liquid samples can provide valuable information. in case, parallel acquisition of thousands of spectra can used compare differences between samples, rather more common implementation of exploring spatial heterogeneity within single sample.

similarly, there no benefit in imaging homogeneous sample, single point spectrometer generate same spectral information. of course definition of homogeneity dependent on spatial resolution of imaging system employed. mir imaging, wavelengths span 3-10 micrometres, objects on order of 5 micrometres may theoretically resolved. sampled areas limited current experimental implementations because illumination provided interferometer. raman imaging may able resolve particles less 1 micrometre in size, sample area can illuminated severely limited. raman imaging, considered impractical image large areas and, consequently, large samples. ft-nir chemical/hyperspectral imaging resolves larger objects (>10 micrometres), , better suited large samples because illumination sources readily available. however, ft-nir microspectroscopy reported capable of 1.2 micron (micrometer) resolution in biological samples furthermore, two-photon excitation fcs experiments reported have attained 15 nanometer resolution on biomembrane thin films special coincidence photon-counting setup.

detection limit

the concept of detection limit chemical imaging quite different bulk spectroscopy, depends on sample itself. because bulk spectrum represents average of materials present, spectral signatures of trace components overwhelmed dilution. in imaging however, each pixel has corresponding spectrum. if physical size of trace contaminant on order of pixel size imaged on sample, spectral signature detectable. if however, trace component dispersed homogeneously (relative pixel image size) throughout sample, not detectable. therefore, detection limits of chemical imaging techniques influenced particle size, chemical , spatial heterogeneity of sample, , spatial resolution of image.

data analysis

data analysis methods chemical imaging data sets typically employ mathematical algorithms common single point spectroscopy or image analysis. reasoning spectrum acquired each detector equivalent single point spectrum; therefore pre-processing, chemometrics , pattern recognition techniques utilized similar goal separate chemical , physical effects , perform qualitative or quantitative characterization of individual sample components. in spatial dimension, each chemical image equivalent digital image , standard image analysis , robust statistical analysis can used feature extraction.

software

fecom object learning software (ols), industrial in-line hyperspectral feature processing
umbio evince image, multivariate hyperspectral image analysis
perception system; in-line hyperspectral imaging industry




^ cite error: named reference near infrared microspectroscopy 2004. pp.241-273 invoked never defined (see page).
^ fecom - object imaging systems
^ perception park - industrial hyperspectral imaging