“Humans are not perfect. They can get distracted and sometimes miss out on subtle speech cues and warning signs. Unfortunately, there is no objective blood test for mental health” - Brita Elvevåg, Cognitive Neuroscientist at the University of Tromsø, Norway

As we know, our emotional, psychological, and social well-being affect the way we think, feel, and act, and are attributed to mental health. It is important over the entire course of our life, right from childhood, and has a bearing on how we cope with stress, relate to others, and make choices. It is very hard not to lose emotional balance, especially during these tough pandemic days that we are all currently wading through. We are always at risk of encountering potential threats to our mental well-being, regardless of what stage of life we’re all in.

At a global policy level, WHO's Global Plan of Action on Workers Health (2008-2017) and Mental Health Action Plan (2013-2030) outline relevant principles, objectives and implementation strategies to promote good mental health at the workplace.

If we look at the statistics, as of 2018,  anxiety affects about 284 million people in the world. There is no doubt that the ongoing pandemic is also contributing to these numbers and the statistics for the current year may look all the more dismal. The World Health Organization (WHO) has also reported that about 25% of the population in Europe experience depression or anxiety every year.

What kind of relationship can there be between mental health and privacy?

Opening up about depression to friends, family, colleagues, and medical professionals can be an important step to start treatment and receiving support. On the other hand, information that discloses when exactly someone is feeling anxious and possible causes of anxiety can be misused to target people when they are the most vulnerable, especially when combined with other data about their interests and habits. Lots of people search for depression and remedial measures online to seek help for themselves or to support others. According to a study conducted by Privacy International, some websites gathering data on depression through surveys share answers and results with third parties. This is a major privacy concern in mental health. Another part is related to advances in mental health-oriented artificial intelligence. Today, as a result of digitization, mental health records and psychotherapy notes are often available as readily accessible health records. This threatens the privacy and confidentiality of patients with clinical depression and related ailments.

Our vision for the project

This project will help us  explore the existing perspective of privacy in  general healthcare  including mental health and benchmark them. It is aimed at developing a privacy-preserving model that achieves our objective: “PPAI for Mental Health”.

Expected milestones

  • Literature review to study existing privacy-preserving data analysis approaches.
  • Investigating the availability of data sources in the realm of mental health.
  • A privacy-preserving data analysis model by benchmarking against existing models.
  • Feasibility research about which libraries from the OpenMined ecosystem  should be used in the model.
  • Open-sourcing the project.
  • Demo.
  • Publish our results!

Join us!

We'll be happy to have you with us in this project that we are excited and passionate to start.

If you aren’t yet a member of the OM Slack workspace, you can join here.

Apply for the project here

💐Thanks to Bala Priya for her editorial review and encouragement.

💐Thanks to Emma Bluemke, Georgios Kaissis and Nahua Kang for their support and encouragement.

References we are inspired by:

[1] Su, C., Xu, Z., Pathak, J. et al.(2020), Deep learning in mental health outcome research: a scoping review.

[2] Miles T (UROC Mentored by Syed Hafiz), Privacy Preserving Detection of Depression from Speech Data.

[3] Wen D, Wei Z, Zhou Y, Li G, Zhang X and Han W (2018), Deep Learning Methods to Process fMRI Data and Their Application in the Diagnosis of Cognitive Impairment: A Brief Overview and Our Opinion.

[4] Qiu H, Qiu M, Liu M, & Memmi G (2019), Privacy-preserving Health Data Sharing for Medical Cyber-Physical Systems.

[5] National Network of Public Health Institutes. Data Governance for Children’s Mental Health Surveillance: What is It and Why Does It Matter?

[6] OECD, Health Data Governance: Privacy, Monitoring and Research - Policy Brief.