Creating digital endpoints for depression and anxiety using the B-Healthy Platform
There are many reasons why as someone engaged in clinical trials of mental health you’d want access to more objective, more robust and more accurate measures of the severity of depression and anxiety.
Early detection of these conditions is super important as it leads to much better patient outcomes through cheaper treatments delivered with less time involvement of highly trained clinicians. Accurate monitoring of the effect of established treatments will get the right treatment to the right person faster than ever. And of course, having more accurate objective measures means greater statistical significance for your clinical studies, which means you either need fewer participants or can be more certain of the outcomes of your studies.
So what’s wrong with the current way of measuring depression and anxiety in clinical studies?
Well, the current measurement of depression primarily relies on self-report questionnaires, with all the well-documented issues that come with that, or on very costly and time-consuming clinician-administered clinical interviews, which prevent running a study at scale and remain subjective to a degree.
Focusing on the most used measurement technique of self-reporting, the downsides of the use of these questionnaires can be boiled down to the following five problems:
Questionnaires are dull, which in turn hinders completion rates.
If you want a particular outcome it’s obvious how to complete the questionnaire - e.g. high scores if you want treatment, low scores if you don’t want the reader to think you’ve got a mental health problem.
Questionnaire fatigue is a real issue when used for treatment monitoring. Completing the same questionnaire every week, or even daily for an intense treatment, causes people to become blind to the questions, relying on memory resulting in similar scores as last time rather than current introspection.
The questionnaires are inherently subjective and open to some interpretation so individuals will complete them with a certain amount of innate bias which is different for every person, whereas an objective score would report the same objective outcome for every individual.
The way the questionnaires are defined means that there is no benefit in measuring more frequently than once a week which is a missed opportunity.
These are really quite damning problems.
What you really want is to replace these questionnaires with an objective measure that can be frictionlessly administered at low cost. Unfortunately, there is no real gold standard for an objective measure. There is no blood test for depression today, and even brain scans can’t diagnose it.
There is strong evidence though that automatic face and voice analysis technology can achieve the same levels of accuracy or even better than self-reports. At BLUESKEYE we have developed a product, TrueBlue which aims to do just that. Our work to date suggests we are succeeding; correlations between our TrueBlue score and PHQ9 are higher than those between PHQ9 and Hamilton-D, for example.
Changing a lifetime of measuring depression with PHQ9 will be difficult, and this is the challenge that BLUESKEYE has taken on. We are just about to begin, close to our home base in Nottinghamshire in the UK, a clinical trial of the safety and efficacy of our TrueBlue product with 125 pregnant and recently postpartum mums in conjunction with the UK’s National Health Service (NHS).
Getting such measures adopted by the healthcare system will result in serious improvements and cost savings and is worth 100x the effort and investment at the least.
Enter digital endpoints.
A good route for wide acceptance and adoption of a new digital endpoint for depression, such as the TrueBlue score, is to first get the TrueBlue score formally accepted as a digital endpoint by the FDA and then adopted in a few clinical trials that measure depression and anxiety.
To that goal, we have created the B-Healthy Platform.
BLUESKEYE AI’s B-Healthy Platform makes it really easy for you to deploy a clinical trial that includes face and voice behavioural biomarkers. Here are the steps you take:
Select what digital biomarker models (endpoints) you want to measure for your study, either based on our documentation or through consultation with our experts.
Together with our UI/UX experts, choose what interactive task best suits your demographic, your protocol, and the endpoints selected in step 1.
Put the app together by configuring the tasks and models together with any other patient reported outcomes you want to measure. Participant information, consent, and notifications can all be configured by you.
The app is available on all major app stores. Your participants can download it and using an activation code activate it.
All endpoint and stage-one output data is stored anonymously on a clinical grade backend ready for you to do analysis on. Or, get our team of world-leading machine learning and data scientists do it for you.
If you get the results you desired, you can then chose to integrate this into a companion app to measure your condition routinely, and off you go!
The B-Healthy Platform provides you with configurable mobile phone modules that you can use to create a data collection app by taking a pick-and-mix approach. By selecting from a set of configurable modules from a library our team can create an app for you in a matter of days. Example modules are participant information and consent forms, video or audio stimuli, and data feedback screens. Most importantly, you can select the interactive task for the participant to perform whilst being recorded. Existing modules in the library include a journalling task, an instruction-following task, a read-out-loud task, an object-tracking task, and others. If there’s a task or a module that you need and that doesn’t exist in the library yet, speak to the team to see if they can fast-track that, we’re always keen to learn what our customers need!
Private, Secure and Fair
Applying our unique two-stage approach to medically relevant behaviour analysis, the original face and voice data is turned into a couple of hundred anonymous parameters for every video frame, that together describe the behaviour. These parameters include things like facial muscle actions, gaze direction, or the tone of voice. These are then sent over a secure, encrypted channel to a secure backend where the anonymous data remains encrypted at rest.
You as the clinical trial manager can then inspect this data, run data science projects on it, or combine it with other anonymised data that you have available through back-to-back server integration for example. All your digital biomarker research can thus be done on entirely private data. A very useful side-effect of BLUESKEYE’s two-stage approach is that all appearance-based biases and invariances are solved at the first stage - it simply HAS to because the image and the audio waveform are discarded afterwards. BLUESKEYE has been proven to be unbiased to apparent gender, age, and ethnicity at this first stage!
BLUESKEYE' s machine learning models have been proven to be unbiased to apparent gender, age, and ethnicity at this first stage!
The B-Healthy Platform is already used by a number of pharmaceutical companies in their studies and is now being rolled out to the wider clinical trials audience. As an ethical AI company, we are well underway with being compliant with the EU AI Act and other AI regulations and you can safely use our services now and in the future!
Proving the value of a digital endpoint
BLUESKEYE AI is incredibly fortunate to be working with DATAcc by the Digital Medicine Society (DiMe) to create tools for the business case for digital endpoints in general (not just for the TrueBlue score), in terms of return on investment from pharmaceutical companies running clinical trials. Together with over 30 pharmaceutical, med-tech, regulatory and academic bodies we will be mapping the stakeholder landscape and creating actionable, publicly available tools that aim to literally allow you to calculate the return on investment in digital endpoints to convince your internal budget holders to adopt them!
If you’re interested in helping create an objective score for depression and anxiety measurement, get it approved as a digital endpoint and widely adopted by the healthcare profession, get involved with DATAcc!