See our published paper 
Facial nerve palsy (FNP) is a neurological condition which adversely affects facial muscles. The absence of a universal assessment tool to assess and grade FNP has resulted in clinicians relying on subjective assessment methods. Existing assessment tools are often manual and time-consuming. This often creates challenges for clinicians to make timely diagnosis and treatment of FNP.
OpenFAS, a collaboration between students at UNSW and clinicians at the Chris O’Brien Lifehouse, is a prototype utilising cheap red-green-blue depth (RGB-D) sensors for automated objective assessment of FNP. The initial clinical results of the system conducted on 16 patients with FNP and age and gender-matched control group are encouraging but not ready for clinical use. The landmarking model was relatively stable, with 88% of all landmarks within the detection threshold. OpenFAS shows a medium to strong correlation with the Sunnybrook facial grading system for dynamic and static analysis but only little to no correlation for synkinesis analysis. Usability results show that this tool is appropriate for clinical use and that most patients are happy to use the tool at all relevant consultations.