09 January 2021 5 951 Report

Dear Colleagues, our imaging facility is entering the field of Raman micro-spectroscopy and we've been approached by a group who want to detect the presence of certain compounds (about 10 different ones) in environmental samples. We can get the spectra of the pure compounds of interest but the samples may possibly contain nearly any compound on Earth, so I imagine the spectra could be pretty messy. I'm wondering what is the best way to find out whether the sample contains any of the compounds of interest. Somebody suggested neuronal networks since we are planning to use that approach for another project. But that project is rather different as we expect only entities with known spectral signatures to be present in the samples, so there is a rather straightforward way hw to train the network. I'm not so sure about the practicability of this approach in the case of the environmental samples and I'm thinking there may be other approaches commonly used for similar types of problems. Any suggestions, links to relevant literature or relevant software are appreciated.

Similar questions and discussions