Software to Measure Facial Symmetry and Predict Dosages and Outcomes of Cosmetic Treatments Such as Botox, Dysport, or Fillers from Video


FACIAL PALSY APPLICATION: Facial palsy is a condition which affects the facial nerve, one of twelve cranial nerves. Its main function is to control the muscles of facial expression; all 17 muscles. This involves the ability to express emotion though controlling the position of the mouth, the eyebrow, nostrils, eye closure, drooling, speech and dental hygiene. A spontaneous facial palsy can present at any age but is most frequently seen at age 20-50 years, affecting both sexes equally. Incidence is around 30 cases per 100,000 per year, and is slightly higher in pregnant women (45per 100,000) and an incidence of 1.8 per 100,000 of children are born with facial palsy.

Paralysis of the muscles which control the movement of the mouth have many consequences. A patient may drool and have difficulty eating or have difficulty moving the affected side of the mouth. When a patient tries to smile, only the normal side of the mouth will move. Speech can also be impaired.

The effect facial palsy has on the upper face largely includes the eye and the eyebrow. If they are unable to close the eye, eventually the eyeball can become dry and cause an ulcer to develop which can lead to problems with a patient’s vision. A facial palsy patient may also lose the ability to raise their eyebrow; this leads to an exaggeration of the asymmetry during facial expression with one side of the forehead having no wrinkles in comparison to the other. Additionally, the profound effect that facial palsy can have on an individual’s psychology is untold.

At present, there are more than 26 scoring systems used to assess either the recovery of the facial nerve or results of surgery. Nevertheless, there continues to be a lack of a universally accepted assessment tool. The difficulty lies in developing an objective, easily reproducible grading system that is able to assess treatment outcomes in a clinical setting and at the same time have significant rigor to be applied to research. While some automated systems have already been developed and validated, they are largely 2-D renderings and capture only static images - failing to incorporate the dynamism of facial movements into their analysis. Additionally, those systems utilizing 3-D images and continued videography are either prohibitively expensive or so burdensome as to effectively render them clinically useless.

COSMETIC APPLICATION: In 2020, Botox became the most popular non‑invasive cosmetic procedure, with 4.4 M procedures done in the US in 2019 according to the American Society of Plastic Surgeons. Non‑surgical cosmetic procedures (neurotoxins such as Botox and dermal‑fillers) are becoming increasingly popular with 5.5 M total procedures performed in 2021. The global non-invasive aesthetic treatment market size was valued at USD 53.8 billion in 2021 and is projected to expand at a compound annual growth rate (CAGR) of 15.2% from 2022 to 2030. While non‑surgical procedures grow in popularity, presently there is no way for clinicians to reliably simulate the response of dosages or volumes of neurotoxins or fillers.

Technology Overview

Researchers at the University of Rochester have developed the Automated Facial Symmetry Index (AFSI) which removes user bias by by reporting outcomes objectively. The importance of objective reporting is two-fold: firstly, to ensure the standardized reporting of results between different groups and secondly, to allow for unbiased evaluation of interventions and new rehabilitation strategies. The AFSI tool is independent of clinician evaluation in determining the primary outcome, symmetry.

In contrast to current outcome measurement tools, the AFSI tracks the overall amount of facial movement in the three discreet regions: the forehead, the eyes and the mid-face. As yet this approach has not been available and would offer superior unbiased data when reporting outcomes.


  • Increases sensitivity and reduces bias compared to clinician assessment
  • Easier and cheaper way to use 3‑D images and videography
  • Easily reproducible grading system
  • Identifies functioning muscle groups that can be strengthened for therapy
  • Allows for a more universal and standardized way to report outcome data


  • For rehabilitation and directed physical therapy for stroke patients
  • Detect subtle changes in the facial nerve from Bell's Palsy and help guide therapy by identifying functioning muscle groups that can be strengthened
  • Also applicable to other chronic disabilities causing abnormalities with facial tone such as stroke patients, chronic facial paralysis, and cerebral palsy patients
  • Outcomes reporting for facial paralysis surgery
  • Potential use for non‑invasive cosmetic treatments involving neurotoxins (Botox, Dysport) or fillers – the purpose would be to simulate the response for dosage/volume of neurotoxin/filler used to offer a bespoke patient experience
URV Reference Number: 6-21079
Patent Information:
Computer Software
For Information, Contact:
Joynita Sur (URV)
Licensing Manager
University of Rochester
Jonathan Leckenby
Andreas Savakis