Facial Recognition Diagnosis Wiki

This wiki was created as a documentation source for team Facial Metrics at Healthcare's Grand Hackfest, March 2014, in Cambridge, MA.

What is the app?
The app should be able to use extant face recognition software to provide discrete measurements of facial characteristics. These characteristics can be indicative of rare diseases.

The app should differentiate Waardenburg Syndrome Types I and II and Kabuki Syndrome from normal ratio parameters captured from the Recognyz analysis.

The app could convert a phenotype or a picture of a phenotype to a genetic variance, which could otherwise only be determined by a gene test. This app is most likely the first app, providing a diagnositic tool for guiding providers.

Who is it for?

 * Patients: Gives patient/caregiver guidance about potential disease states; measurement tool
 * PCP: Provide educated guesswork to narrow further study and reduce diagnosis time

Measurements
Preliminary tests of conversions from pixel to a cm measurement were successful. The method of calibration is documented here

Disease States
Disease states affecting eye distance measurements for use presentation and initial demo


 * Waardenburg syndrome
 * Kabuki Syndrome
 * Marfan Syndrome

Outcomes
Less expensive, quicker option for difficult to diagnosis disease states.


 * 1) Diagnosing a disease
 * 2) Directing genetic testing

If a doctor only gets one test reimbursed (current standard), he/she can test for the right disease state. Therefore, directing the test is very important.

App should differentiate Waardenburg Syndrome Type I and II and Kabuki syndrome from norma by ratio parameters in the image. Therefore taking a picture of a child with any one of the diseases would be great.

The app could also be used as a web service for parents to send in pictures of their children to get a list of potential diagnoses.

Presentation

 * Presentation Chart
 * https://docs.google.com/presentation/d/1ohPeq3P36wza2WQib5tRMCYH993idMPOXPpGiQrqIkQ/edit?usp=sharing

Business Model
Business Model

Team Members
Team Members