| Early Detection of Cancer Using LIBS Early detection of many forms of cancer has proven to be vital in reducing the mortality rates of many forms of cancer. From a physics point of view this is an interesting confluence of fundamental science and application. While we are helping to solve one of our great societal problems, we also investigate micro/nanoparticle interactions in laser–induced plasma during tagging procedures, laser-plasma-material interactions while investigating the substrate dependence of our results, and determine and expand the role of machine learning tools in analysis and detection of spectroscopy for cancer detection. Our work in this area has sought to develop LIBS as a technique to diagnose certain types of cancer in a minimally invasive fashion using bodily fluids such as blood. One such method is to tag specific biomarkers and look for changing in the LIBS signal using a technique call TAG-LIBS. We have previously shown this to be feasible for human blood by tagging the biomarker CA 125 for the detection of epithelial ovarian cancer.  Current efforts have focused on direct interrogations of blood.The figure to the right shows a process we have used to study the detection of melanoma in mice. We investigated blood deposited on several substrate surfaces and investigated discrimination between healthy and cancerous subjects using several machine learning algorithms. Careful selection of the substrate and machine learning algorithm yielded a 97% classification accuracy. Continued interests and further work on this subject involves considering extensions other diseases.This work is driven by past and on-going collaborations with several institutions and research groups including the Memorial Sloan Kettering Cancer Center, Edith Nourse Rogers Memorial Veterans Hospital, and the Fox Chase Cancer Center.ul selection of the substrate and machine learning algorithm yielded a 97% classification accuracy. Continued interests and further work on this subject involves considering extensions other diseases. |