The Future of Preventative Behavioral Healthcare Powered by Machine Learning

Improving preventative mental health care through passive a smartphone application enabled by machine learning

(702 Words)

The Challenges of Addressing Mental Health Disorders

Mental health disorders are on track to be the leading cause of disability by 2020, with an annual global economic impact of $1 trillion.[1] Machine learning technology utilized at Mindstrong Health aims to dramatically impact healthcare providers ability to diagnose and treat these disorders. The current process of diagnosing and treating is ineffective because it requires the patient to break from their normal routine, display symptoms at the time of the assessment, and is difficult to scale due to the requirement of a staffed facility.[2] Mindstrong is addressing mental health needs by using machine learning algorithms to collect and analyze data from patients smartphones to rapidly assess and provide updates to mental health providers. [3] Through early detection and intervention, Mindstrong’s solution reduces healthcare usage, while providing better patient care.

 

Mindstrong Health’s Ground-Breaking Solution

Mindstrong Health focuses on bringing measurement science to mental health services. Its smartphone application passively acquires data by monitoring how a user interacts with their smartphone through swipes, taps, and keystrokes, rather than the content of what they type. This data is encrypted and analyzed remotely using machine learning and compared to Mindstrong’s database to assess a patients mental health. The resulting cognition and emotion bench marking information, referred to as digital biomarkers, are shared with the patient’s medical provider to allow the provider to better serve the patient.  Clinical studies with leading research universities confirmed that Mindstrong’s application was very effective in monitoring patients mental health, specifically for identifying when a patient was depressed.[2]

 

Challenges Faced by Mindstrong Health 

A few critical questions management is focused on addressing over the next few years are: [5]

  1. How and when to market its smartphone application to the general market
  2. How to maintain security of sensitive patient data
  3. How to expand the product beyond initial successes in depression

Mindstrong’s product currently serves high risk depression patients but it is planning to release its application to the public over the next few years. [5] Expanding the population sample will be critical to improving the accuracy of its data set, but questions remain over whether the general population would adopt a passive software solution that constantly evaluates and provides updates to providers of the users mental health conditions. Critical to building broad consumer appeal for its application will be maintaining security of patient data and using the data only for the purposes described to patients. [6] While Mindstrong has been explicit on how it collects, uses and shares patient data, management is evaluating the appropriate long-term strategy.[7] Building upon its initial successes with depression, Mindstrong is also focused on expanding its addressable market by identifying digital biomarkers associated with other mental health disorders such as, psychotic disorders, schizophrenia, and PTSD by conducting clinical studies with universities such as Stanford, the University of Michigan, and Kings College. [4]

 

How Mindstrong Health Can Address these Challenges

To address the identified issues in the short and medium terms, I would suggest the company focus on transparency around product efficacy, accelerate product development, and simplify its message on data security.

  1. Mindstrong can accelerate the general perception of product effectiveness by providing additional transparency on how its patented machine learning algorithm identifies digital biomarkers by sharing software code and data findings with the public.
  2. Important clinical studies in high commercial potential areas such PTSD are not expected to be completed until 2022. Mindstrong could accelerate the trials and protect its market leading position by allocating additional resources to the studies.
  3. If Mindstrong hopes to bring its product to the general market it must simplify its message on how it collects, analyzes, stores, secures and shares patient data. Its website provides this information in a form that is too dense for the general consumer to understand.

 

Open Questions:

  1. What are the implication of our smartphones having the ability to assess our mental health and provide insights and alerts to healthcare providers?
  2. What is the best channel strategy to bring Mindstrong’s consumer facing product to the market to optimize commercial potential? (ie DTC, through health networks, through providers or just to patients at risk of mental illness or healthy patients?)

 

 

[1] Brunier, Alison. “Investing in Treatment for Depression and Anxiety Leads to Fourfold Return.” Investing in Treatment for Depression and Anxiety Leads to Fourfold Return. November 13, 2016. Accessed November 13, 2018. http://www.who.int/en/news-room/detail/13-04-2016-investing-in-treatment-for-depression-and-anxiety-leads-to-fourfold-return.

[2] Dagum, Paul. “Digital Biomarkers of Cognitive Function.” Npj Digital Medicine 1, no. 1 (2018). doi:10.1038/s41746-018-0018-4.

[3] “Science.” Mindstrong Health. March 28, 2018. Accessed November 12, 2018. https://mindstronghealth.com/science/.

[4] “Clinical Programs.” Mindstrong Health. March 28, 2018. Accessed November 12, 2018. https://mindstronghealth.com/clinical-programs/

[5] Metz, Rachel. “The Smartphone App That Can Tell You’re Depressed before You Know It Yourself.” MIT Technology Review. October 30, 2018. Accessed November 13, 2018. https://www.technologyreview.com/s/612266/the-smartphone-app-that-can-tell-youre-depressed-before-you-know-it-yourself/.

[6] Wyman, Oliver. “Cybersecurity: The Hidden Dangers of Healthcare IoT.” August 22, 2016. Accessed November 13, 2018. https://health.oliverwyman.com/2016/08/cybersecurity_theh.html.

[7] “Privacy.” Mindstrong Health. October 24, 2018. Accessed November 12, 2018. https://mindstronghealth.com/privacy/

Previous:

Realizing the Promise of Additive Manufacturing at Boeing

Next:

Envision Energy: Use machine learning technology to achieve an energy efficient and low-carbon future in China

5 thoughts on “The Future of Preventative Behavioral Healthcare Powered by Machine Learning

  1. I think this technology is very interesting for addressing an important issue, but I’d be concerned if a person knowing they have Mindstrong on their phone impacts how they use their phone. Similar to your point that individuals act differently in psychological examinations, would they act differently on their phone to impact a potential diagnosis? If so, this could cause the machine learning technology to believe the wrong data is or is not correlated to mental health issues and misdiagnose individuals – I would be concerned about relying too fully on this data as you could miss important diagnoses and negatively impact individuals’ lives.

    Also, as you mentioned in the case I think there are a lot of data concern and privacy issues that would come along with this technology, but believe the potential upside (if the data can accurately diagnose) is worth these risks if they are handled correctly.

  2. This is a great piece on a very interesting technology that is likely to have a big impact on diagnoses of mental health issues. However, I worry if this technology will be accessible to the people who are most at risk and need it the most. People in lower income classes suffer from more mental health issues than those from upper classes, and may not have the same accessibility to smart phones. According to the Pew Research Center, smartphone penetration in the U.S. is 77% (http://www.pewinternet.org/fact-sheet/mobile/). The 23% of Americans who don’t have smartphones are likely lower class Americans, who need Mindstrong the most.

    I wonder how we can ensure that this technology spreads to areas of the country and to the groups of people that are truly in need.

  3. This is such an interesting piece about an interesting product. I agree with the other commenters concerns about machine learning applications in healthcare. The issues raised there can help address your second question of the best go-to-market strategy for Mindstrong. Namely, the go-to-market strategy will need to address the concerns of data security and the efficacy of the analysis of that data. I think one way to address those challenges is by partnering with a mental healthcare provider. There are over 550K+ providers in the US alone (https://psychcentral.com/lib/mental-health-professionals-us-statistics/) so there is significant opportunity for partnership. Furthermore, partnering with providers could add a sense of legitimiacy in the crowded field of mental health apps (there are over 10K according to this site:https://www.mdedge.com/psychiatry/article/159127/depression/mental-health-apps-what-tell-patients). I also think it helps protect the company as the providers can help mediate the results and treatment protocol. All in all, I think this could be a helpful tool if deployed correctly.

  4. Hi Kentucky Freud Chicken, thank you for this interesting piece on mental health! I also wrote about a similar product, but noticed your essay had a much stronger focus on patient and data privacy. This is an interesting point, as Mindstrong will control effectively clinical data that is derived from people’s iPhones – to what degree to you believe this data should be HIPAA-compliant? What class of privacy should this data fall into and what kind of requirements should be put in place to encrypt it, etc. in order to prevent misaligned incentives for Mindstrong to eventually sell this data to health insurance companies or employers? Additionally, regarding your question around commercialization, what types of players in the healthcare system do you believe is most interested in this type of product? Would you market it to providers, or rather work with the payers (health insurance companies, Medicaid, Medicare, etc.)? I strongly believe it will be the payers that would be most willing to pay for this solution and make it feasible to bring to the broader market, but its the providers whose buy-in is needed to make this product most effective, a challenging obstacle.

  5. A very promising idea in a greatly underinvested area.

    However, the idea is based on an assumption that there are common patterns in the use of digital devices, indicating mental problems. There could be a great variation among this patterns that could create a lot of false positive outcomes. In other words the fast scrolling with a certain angle could be a mental disorder could be caused by a mental disorder for one patient and be just a habit for another.

Leave a comment