PARISS Hyperspectral Imaging Workflow Using Spectral Waveform Cross Correlation Analysis (SWCCA)
Data processing in a low signal to noise (S/N) environment is a challenge for many algorithms, especially when a signal is near buried in noise or background.
Hyperspectral microscopy applications such as nanoparticle characterization, fluorescence labeling, and weak emissions often present low S/N conditions that challenge linear algorithms.
However, low S/N in a complex background is an everyday challenge faced by the seismology, radar and sonar communities.
Their solution? Use cross correlation algorithms (CCA) rather than the traditional linear algorithms used by the imaging communities.
Team LightForm determined that CCA, when applied to hyperspectral imaging data sets, increased sensitivity and reduced acquisition times. (CCA becomes SWCCA when the application presents spectral “waveforms”)
SWCCA is linearity independent working equally well under both linear and non-linear conditions. Therefore, SWCCA is not subject to the errors produced by misapplying linear algorithms to non-linear relationships.
PARISS Software Extracts Signal From Noise Even In Non-Linear Conditions
- Utilizes proprietary algorithms developed in house, based on Spectral Waveform Cross Correlation Analysis (SWCCA)
- Linearity independent SWCCA algorithms accommodate non-linear spectral mixing that normally occurs in biological samples.
- Highly tolerant of low S/N spectra: enables the generation of robust Reference Spectral Libraries (RSL) and correlation between sample spectra and RSL spectra.
- Reference Spectral Libraries that truly represent your samples.
- Enables accurate spectral segmentation.
- Standardized spectra Spectra acquired with PARISS can be standardized in absorption, % transmission/reflection.
- Publication ready spectra can be compared with spectra acquired on any other analytical instrument.
- Powerful topographical mapping: spectra from the FOV that “correlate” with RSL spectra can be pseudo-colored and “painted” onto a gray-scale image, with pixel-perfect accuracy.
PARISS SWCCA Hyperspectral Software Includes:
- Logical operators: “equals” and “not equals” at a user defined threshold. Controls risk of false positives or negatives.
- User created “real life” reference spectral libraries:
- Spectral topographical mapping: Map the location in a FOV of all or some target objects.
- SWCCA algorithms capture natural variations in spectral profile that indicate change such as: pH, ion-concentration, charge, and conformation.
- Measure change over time: PARISS can acquire spectra automatically over a user defined period, then play the results back as a movie.
- Observe raw spectra: all acquisitions can be exported to third party math or imaging programs.
PARISS users: access your active operating manual (password required) here
How PARISS Hyperspectral Microscopy Works