INTERNAL INFORMATION ONLY – NOT FOR EXTERNAL SHARING

Brief:

AHI Labs was requested to provide a direct like-for-like comparative analysis of the main differentiating features of AHI FaceScan offering versus that of our licensed supplier and competitors, to inform the marketing and sales teams.

In summary, AHI’s FaceScan is able to provide clients with an enhanced offering based on the following 5 features:

1. When delivered within the BHA feature, AHI makes use of more data being collected by way of clinically validated assessments (GAD-7¹, PHQ-9² and others). These additional data sets are then applied within a broader scientific and clinical framework to improve the health risk predictive models whilst providing a more comprehensive and holistic view of the individual.  AHI has leveraged a multitude of reputable peer-reviewed scientific research papers.  

2. Within the BHA AHI collects clinically meaningful data by subjecting the user to clinically and scientifically validated measurement protocols. This approach allows us to ensure that we feed the appropriately measured metrics to our downstream health risk models in order to produce meaningful results (e.g. a non-resting blood pressure or not breathing naturally while doing your measurement will lead to false results).

• AHI uses this first scan to help triage people into more in-depth screenings to both broaden and deepen the risk insights per individual either as a singular snapshot or on a continuum over a period of time. The data and measurements of FaceScan is efficiently re-used towards creating these increased risk model insights.
• AHI has a more sophisticated layer of reporting which can offer correlative analysis based on a multitude of parameters (biometric, behavioural, epidemiological heat mapping) that can be customized to the clients’ needs. (We need to check this as I am not sure how or what they do beyond what we experience on the app)
• AHI also offers a level of both clinical and operational consulting to assist our clients in the application of both the technology and the solutions within their environments. (Vlado not sure if they offer this or if they do it at the same level that we do)  

Further work:

We have managed to do some comparative data analysis of their results vs our BHA but more we will, due to more data points, be more confident with the analysis once we re-run the models of just FaceScan with our scientific measurement protocols and in combination with the clinical assessment outputs. The inclusion of the new assessments is essentially being done as part of the MVD version which will give us the first bullet points outcome.  

We would like to get closer to the raw data and features that is obfuscated from us to see how our modelling will be driven differently going forward, I believe we are still mostly reproducing our licensed supplier’s derivative results in our version of FaceScan (in lieu of our protocols and assessments). I recommend we get a view of what is being altered for each output that FaceScan publishes to match against their ouputs.

We have also after an initial scientific literature search found academic groups that have reproduced and even improved on blood pressure measurement by our licensed provider.  We are looking into these papers, whether there are code available, and whether there are licenses that align with commercial interests.  

Risk

There is still a risk that if they run our current FaceScan and our licensed provider back-2-back that not all of the above benefits will be tangibly visible in the results.  

References:

1. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Arch Intern Med. 2006;166(10):1092–1097. doi:10.1001/archinte.166.10.1092

2. Kroenke, K., Spitzer, R.L. & Williams, J.B.W. The PHQ-9. J GEN INTERN MED 16, 606–613 (2001). https://doi.org/10.1046/j.1525-1497.2001.016009606.x