MelaFind® features a hand-held imaging device that emits multiple wavelengths of light to obtain thousands of characteristics of suspicious pigmented skin lesions up to 2.5 mm deep into the skin. The system uses sophisticated, proprietary algorithms to analyze these data against a proprietary database of melanomas and benign lesions. Based on the analysis, the system generates an objective, detailed report with an immediate recommendation of whether the lesion should be biopsied. The key components of the MelaFind® system are listed below:

A hand-held imaging device comprised of several components:

• an illuminator that shines light of 10 different specific wavelengths, including near infra-red bands;
• a lens system composed of nine elements that creates images of the light reflected from the lesions;
• a photon (light) sensor; and
• an image processor employing proprietary algorithms to extract many discrete characteristics or features from the images.

Proprietary MelaFind® database of pigmented skin lesions. The database includes in vivo MelaFind® images and corresponding histological results of approximately 6,000 biopsied lesions from approximately 4,500 patients. EOS believes this is the largest such database in the US and a substantial barrier to competition.

MelaFind® lesion classifiers. The “brain” of the MelaFind® system, these sophisticated mathematical algorithms distinguish melanoma from non-melanoma by comparing the lesion’s features with the characteristics of the benign and malignant lesions stored in EOS’ the proprietary MelaFind® database.  

Lesion classifier development and training is an iterative process involving: (1) selection of the lesion features that provide for optimal lesion discrimination; (2) optimization of the mathematical formulas to differentiate benign lesions from melanoma; and (3) expansion of the size and diversity of our proprietary lesion database. The performance of the lesion classifiers is directly related to the size of the database used in classifier training, as well as the degree to which the training database is representative of the lesions that will be evaluated by MelaFind® in commercial use.

As with many diagnostic systems, the diagnostic performance of MelaFind® is characterized using two measures: (1) sensitivity — the ability to detect disease when it is present; and (2) specificity — the ability to exclude disease when it is not present. Since sensitivity and specificity are typically trade-offs, meaning that as one parameter increases the other decreases, the MelaFind® lesion classifier is developed and trained with the intention that MelaFind® will detect all melanomas in the training data set with the highest possible specificity.

MelaFind ® is currently under clinical investigation and is not yet available for commercial distribution in the United States or any other country. EOS has entered into a binding Protocol Agreement with the US Food and Drug Administration (FDA). EOS believes the presence of the Protocol Agreement, under which EOS and the agency agreed upon the specific parameters and conduct of the pivotal trial in order to establish the safety and effectiveness of MelaFind®, significantly enhances its ability to expedite the FDA approval process.