Fast, large area, multiphoton exoscope (FLAME) for improving early detection of melanoma

Project Summary
Early detection of melanoma is a key factor in improving patient survival and decreasing treatment
costs. The sensitivity of dermoscopy, the standard of care in the diagnosis of melanocytic lesions, was
reported to be highly variable, ranging between 68-96%, depending on the proficiency of the physician and the
stage of the lesion. Low sensitivity reflects high rates of false-negative findings, which delay diagnosis and
treatment. Thus doctors must err on the side of caution, which leads to an excess of unnecessary biopsies and
increased medical costs. Distinguishing cutaneous melanoma from benign melanocytic nevi with high accuracy
based on dermoscopy remains a challenge even when in the hands of expert clinicians since this approach
only offers a two-dimensional image of the lesion's superficial structure. Ultimately, a biopsy is necessary for
definitive diagnosis by the dermatopathologist, but this too may be affected by inter-observer variability,
resulting in discordant conclusions. A study performed at the Melanoma Center, at UCSF estimated that
214,500 to 643,500 cases of melanocytic neoplasms in the United States would be diagnosed differently by
another dermatopathologist, annually, which has significant consequences for the patient regardless of the
nature of the lesion. We propose to develop and clinically evaluate a fast, large area multiphoton exoscope
(FLAME) as a tool for non-invasive imaging and early detection of melanoma in order to reduce false positives
and false negatives in both dermoscopy and histopathology. Multiphoton microscopy (MPM) is a nonlinear
optical imaging technique that provides unique structural and molecular contrast based on endogenous signals
such as second harmonic generation from collagen and two-photon excited fluorescence from NAD(P)H/FAD+,
keratin, melanin and elastin fibers. In preliminary studies, we demonstrated that macroscopic areas of skin
(cm2 scale) could be mapped out with microscopic resolution within ~2 minutes by combining optical and
mechanical scanning mechanisms with deep learning image restoration. As required by PAR-20-155 our
academic-industrial partnership will deliver a powerful MPM imaging tool to clinicians for non-invasive, real-
time quantitative assessment at the bedside that would not require specialized training. Our proposed
application is for early diagnosis of melanoma, but the approach will have wider impact, for rapid, in vivo
characterization of cellular morphologic and metabolic imaging endpoints in patients. Our specifics aims are:
(1) to develop FLAME, a compact, portable MPM prototype system for rapid, depth-resolved in vivo imaging of
skin, over macroscopic areas (cm2-scale) with microscopic resolution and enhanced molecular contrast; (2) to
implement safety features and demonstrate the technical feasibility; (3) to test the performance of FLAME by
evaluating its ability to provide in vivo quantitative optical endpoints with sufficiently high predictive power to
reliably distinguish benign from early melanoma lesions. We are a multi-disciplinary team of investigators from
UC Irvine, Vidrio Technologies, LLC and Tufts University with 3 to 8 years record of collaboration.

Grant Number: 4R01CA259019-04
NIH Institute/Center: NIH
Principal Investigator: Mihaela Balu