Spectrally Guided Mohs Surgery

The current standard-of-care in the identification of skin cancer is visual inspection followed by biopsy and histopathology of suspicious skin sites. Since a physician is required to perform this biopsy, there is often a delay in diagnosis, resulting in deeper, more aggressive tumors and increased mortality from malignant melanoma (MM). Therefore, a non-invasive method to inspect these lesions would be of great clinical importance.

An initial prototype of a noninvasive diagnostic device was developed based on optical spectroscopy and completed a clinical study in 76 patients that demonstrated high diagnostic accuracy for the detection of skin cancer (IRB # CR-10-004). This initial prototype consisted of two separate devices and probes: one to collect Raman spectra (RS) and the other to collect diffuse reflectance and laser induced fluorescence spectra (DRS+LIFS). type, but a combination of modalities gave the best diagnostic performance for all types of skin cancer.

The addition of Raman spectroscopy improved diagnostic performance for both melanoma and non-melanoma skin cancer. However, the operation of the integrated systems was still conducted via two optical fiber probes (the first one for fluorescence and reflectance spectroscopy, the second one for Raman spectroscopy). The need to take measurements of the same lesion using two probes increased acquisition time, and the possible sampling site error. Recently, a device was developed that combined fiber optic probe that is capable for spectral acquisition of Raman, white light reflectance and laser induced fluorescence spectroscopy. Using this probe, acquisition time and sampling site error should be reduced. There is no significant difference in terms of performance between the previous two probes and the new probe.

Models have been developed to analyze reflectance and fluorescence spectroscopy data. In order to interpret Raman spectroscopy data in physiologically relevant parameters, a biophysical model needs to be developed. Similar models have been developed by other research groups for other types of tissue.

This study proposes to use the new technique of biophysical modeling to analyze our Raman spectra. At the core of the technique is the measurement of a set of "basis spectra" which are fit to the data using ordinary least-squares. Recently, biophysical models have been developed for atherosclerosis and breast cancer with very impressive diagnostic results, achieving 94% sensitivity and 96% specificity for breast cancer and 94% accuracy for atherosclerosis disease classification.