Multimodal Spectroscopy (MMS) for in Vivo Noninvasive Assessment of Skin

The goal of this research program is to develop a simple, noninvasive diagnostic device for assessing skin pathology without the need for a biopsy. The device being studied is a single system capable of collecting three spectroscopy measurements (Raman, diffuse reflectance and laser induced fluorescence spectroscopy) from skin lesion sites. In order to accomplish this objective, the investigators propose to: 1) develop biophysical model for Raman spectroscopy of skin cancers, 2) conduct a clinical pilot study to collect MMS data from a minimum of 250 patients to determine the diagnostic accuracy (sensitivity and specificity) of MMS for diagnosing skin malignancy.

To develop a biophysical model the investigators will recruit patients based on their known histopathology diagnoses, from each of 6 skin-cancer categories: basal cell carcinoma, squamous cell carcinoma, melanoma, atypical nevus, actinic keratosis, and benign skin. Patients with non-cancer diagnosis such as psoriasis, eczema / dermatitis, lichen planus or lupus will also be recruited. Recruitment will include 5 patients from each of these categories, giving a total of 20 patients. Measurements for the biophysical models are conducted on skin tissues that have been excised from the patients. Furthermore, these skin tissues are excised tissues that under standard screening procedure will be biopsied by the dermatologist. Therefore, patients are not undergoing extra \& unnecessary biopsies, and will not be exposed to any risk involved with the microspectroscopy measurements. To determine the diagnostic accuracy of MMS, MMS data will be collected from six clinical groups, each with a preoperative diagnosis: 1) malignant melanoma (MM), 2) basal cell carcinoma (BCC), 3) squamous cell carcinoma (SCC), 4) pre-cancerous lesions (AK), 5) benign or atypical nevi, or 6) other lesions or conditions (rare skin cancers such as merkel cell carcinoma, and inflammatory conditions such as psoriasis). The aim is to collect MMS data from a minimum of 240 patients (40 patients for each of the 6 groups), which will be split to training and validation datasets. Since several biopsies are required to diagnose a single skin cancer, it is anticipated that the actual number of pre-cancerous, benign and inflammatory lesions sampled will be much higher than our target total of 240, to be around 300. We will also collect data from non-skin cancer skin conditions: 1) psoriasis, 2) eczema / dermatitis, 3) lichen planus and 4) lupus. For the non-skin-cancer part of the study, we will enroll 20 patients from each of the categories, with a total target of 80 patients.