If you log into any social media channel right now, you will notice various quotes and conversation threads highlighting, even empathising with those suffering from depression or repressed mental health issues. While one can’t be sure for how long these conversations are going to continue with the same vigour, the fact that people are coming out and acknowledging the mental health problem is overall a step in the right direction.
In 2019, WHO had estimated that 20% of India’s population will suffer from a mental health issue this year. Given that WHO had not accounted for the isolation that COVID-19 would bring, this estimation may be far from the grave reality today.
Before we explore how mental health problems are diagnosed and treated, and why data science can be its biggest ally in treatment, let us first look at the challenge that the doctors treating the patients are facing, today.
When opening up about her struggle with depression, famous Hollywood celebrity, Anne Hathaway shared with the world, “I disliked myself so intensely. It was just a mindset. I didn’t know how to love myself. I didn’t know how to love anybody.” She isn’t the only one. While people suffering from mental health issues across the globe struggle to not only recognise the symptoms, and truly acknowledge the gravity of the issue at hand, doctors often struggle with choosing the right course of treatment.
Often than not, this challenge of not having a defined route of treatment results in adopting a trial and error route, which further prolongs a patient’s treatment without an assurity of cure. This can be both harrowing and painful for the patient and incredibly frustrating for the doctors themselves.
To bring structure to the current course of treatment and possibly, even narrow the gap between the patient’s diagnosis and final course of treatment, healthcare professionals require a methodology that can help them identify patterns that are otherwise difficult to detect. This route of treatment can be found with the help of data science.
Moving forward with data science for mental healthcare
At present, this approach is in its nascent stage of adoption. While healthcare professionals are slowly but steadily warming up to this partnership, the cynics are doubtful that analysing data can help treat complex mental health issues. However, the majority believe this is the step in the right direction.
In fact, many researchers believe that big data can even put an end to the current course of treatment, ie. trial and error. While there isn’t a single source of information currently available, these researchers feel that there is a huge scope of study in using data science in the field of mental healthcare
Let us explore how data scientists can help resolve the current dilemma.
Data science uses scientific methods and processes to derive insights from both structured and unstructured data. Data scientists can apply their knowledge and skills to create models that will help link complex data sources currently available in mental healthcare to arrive at constructive conclusions.
If these models can be created for mental healthcare, it will enable researchers to extract the risk factors for mental illness, which is currently unavailable. To be able to identify the risk factors can further pave the way for doctors to give the right diagnosis and treatment, map mental health improvement and even study how the illness can be prevented from occurring in the first place.
Big data and Data science have the potential to create personalised care for patients suffering from mental illness, the lack of which is currently causing a wide gap between those suffering and those who have recovered/ or are recovering.
A closer look at some success stories
While it seems that the role of data science in developing line of treatments for mental health patients may seem too good to be true, many organisations have already reaped its benefit by acting on these theories. Data scientists at the Crisis Text Line, a global not-for-profit organisation based out of the United States which has a text messaging-based crisis counseling hotline, use machine learning to analyse words and emojis in text messages that they receive.
Machine Learning helps them extract those words and emojis that can indicate a higher level of suicidal tendencies or even self-harm. This further helps them prioritise the patients by identifying who needs the maximum help and thus, should be addressed first. To arrive at this stage, data scientists at Crisis Text Lines analysed over 30 million texts exchanged with the users.
The analysis even threw up other insights like maximum anxiety related cases were received on Wednesday and self-harming cases arise maximum at the darkest hours of the night. In a country that doesn’t have a single test or panel of tests to identify a possible suicide crisis, this utilisation of AI by data scientists to prevent suicide can help save many lives, today.
Another example that comes to mind is ‘Ellie’. Back in 2017, a virtual therapist named Ellie was put to interact and analyse the behaviour of soldiers who had recently returned from Afghanistan. Surprisingly, Ellie was successful in discovering more signs of Post Traumatic Stress Disorder (PTSD) symptoms in the soldiers than the Post-Deployment Health Assessment personnel themselves.
The list of examples showcasing how data science and machine learning has helped gain exemplary results in the field of mental healthcare are many. However, these examples continue to exist in pockets. The good news is that countries are at least opening up to its possibilities. From the World Well Being Project to an app which monitors nocturnal patterns to help university students combat depression, this data science bridge is perhaps the need of the hour.
However, before signing off, it would also be important to note that mental health is an extremely personal line of healthcare service. Given that the issue itself was stigmatised for many years, and continues to be so for many regions, there is a serious concern of privacy. The question of ethics arises when we bring in the possibility of algorithms providing clinical diagnoses or recommendations.
Many mental healthcare smartphone applications today lack an “underlying evidence base, a lack of scientific credibility, and limited clinical effectiveness.” So, while the future of data science in mental healthcare is promising, we need to tread carefully with the right infrastructure and laws in place that protect the patients without compromising their privacy for the benefit of a potential boon.
If you are curious to learn about data science, check out IIIT-B & upGrad’s PG Diploma in Data Science which is created for working professionals and offers 10+ case studies & projects, practical hands-on workshops, mentorship with industry experts, 1-on-1 with industry mentors, 400+ hours of learning and job assistance with top firms.