We use femtosecond light pulses sent though a microscope to excite and read vibrational states of molecules, and spatially map their location within a sample. The vibrational signals give us unique identifying fingerprints of those molecules. Many of the chemical species we resolve cannot be labeled, and so are not typically captured in micrographs. Ability to spatially mapping these and other species gives us unique insights into functional states of the systems we study. Current efforts under this project include:
- Development of faster, simpler approaches to acquiring BCARS micrographs. These efforts involve manipulating ultrafast laser pulses in a microscope and retrieving Raman spectra from the resultant signal.
- Application of BCARS microscopy to biology and biotechnology. This work allows us to leverage our unique microscopy tool, and frequently involves collaborators. We have contributed to several notable discoveries in biological systems, and have more manuscripts along these lines currently in preparation.
- Machine learning and application of BCARS to medical diagnostics, particularly histopathology. Because BCARS provides a comprehensive representation of chemical species in a sample, it is a very powerful nascent tool for detecting presence of disease in human tissue. These tissues are quite complex, and the Raman spectra we obtain from them reflect that complexity. Thus, we use various machine learning tools, including multivariate analysis and neural networks, to extract information from our spectral images.
Camp, C. H., Jr., Lee, Y. J. & Cicerone, M. T. Quantitative, comparable coherent anti-Stokes Raman scattering (CARS) spectroscopy: correcting errors in phase retrieval. J. Raman Spectrosc. 47, 408–415 (2015).