Why? (The problem of mental health)
The rapid increase in cases for mental health issues during and post lockdown period clearly shows an urgency to find a permanent solution for this problem. Cases of depression, stress, anxiety and panic attacks suddenly witnessed a huge rise from 3-5% before 2020 to 21-27% in 2021. Loneliness, lack of communication with family and friends can be some of the reasons but according to various studies, social media and news channels played a major role. People became mentally unstable and relatively unavailable to their near and dear ones.
Technology playing a crucial role in hampering mental health unintentionally. So instead of blaming technology, can we use it for solving the problem? Definitely yes, so here I am proposing a solution using Artificial Intelligence and Machine Learning to cope up the mental health issues and concerns and help people.
The idea is to quantify the mental health. So just like any other medical test for example blood test which give a proper report having the readings of all the elements in the blood along with their levels. We need a proper quantifying mechanism to examine the status of mental health based on various real-time aspects of life in accordance with the emotional intelligence. Like when you know that you have low hemoglobin, then only you can improve the levels. Similarly, if the mental health quantification can be able to give the exact analysis of the brain, thoughts, intensity, stress and hormones like dopamine or serotonin, we can achieve the way through which a mind can be programmed into a peaceful one.
How can we use technology to deal with mental health?
I have divided the entire idea into 2 parts:
- Quantification of mental health
- Treatment using quantification of mental health
Quantification: The scoring system
The idea is to first create a list of the all-possible factors helpful to determine the current situation of mental health. The list must contain 400+ factors decided after taking into consideration the opinions of the best mental health specialists from all different counters. This is because every country has their own culture, tradition, values and lifestyle and hence the impact may differ in patients too. Some of these factors can be:
- The search pattern of the person: The search pattern means the way a person uses a search engine and what type of things he/she is searching for. It may be biased as a person might be a pharmaceutical company employer searching for various types of medicines but the mechanism may give a low score to him because it may take him as an unhealthy person. So the accuracy and precision of the algorithm must be maintained to deal with these scenarios and other 399+ factors will play their role in such cases to predict the exact score.
- The static coordinates: So it is been said that a person suffering from such issues don’t really like to leave his room. Like the way, google track our location for directions by taking each step into counts, we can observe the frequency of change in coordinates daily, monthly or for different intervals. We can determine if the person even changes the room, go to terrace or not, steps out of the house for any purpose or not apart from office work or not.
- Cookies of the devices: The cookies contain the metadata of users which helps in determining the interests and needs. Even companies like Google, Facebook use these cookies only to target users for their advertisements. The same method can be used to know the mindset of any user.
- Instagram scroll: Like Instagram now determines the taste of user by just observing 10-15 minutes. Analyzing the likes and retention time of users when they scroll, Instagram gives similar results only. So, these observations and algorithms can be refined to create intelligent metrics based on the user’s interest and mindset situation.
- Level of light in a room: Like its been said that the patients or people suffering from these issues like darkness more, so if we can monitor the brightness of the room and average intensity, the metrics can also play an important role in the algorithm to determine the mental health score.
- Voice frequencies and variation: A sad person has a low frequency while an excited one has a higher one. So if voice variation frequencies and time duration can be measured and calculated, the heaviness of thoughts can be determined.
Other factors like music taste, words they use while on calls, messages or emails, how they react etc. collectively can contribute in the quantification process.
Once the score out of 10 is out, let us categorize it into three types;
- NEED SERIOUS ATTENTION (<3.5/10)
- AVERAGE STATE (6/10)
- HEALTHY STATE (>7.5/10)
Treatment of mental health issues
The score will also give the type of mental health issue like if it is due to breakup, diseases, fear of failure, death of near and dear one or like what happened in covid due to the news and environment during pandemic. The algorithm will also categorize the doctors or specialist in panel according to their success ratio for the type of cases they deal with. For example, if a mental health expert A got 90% success in treating people from fear of failures, then the affected person will get him in the recommendation.
Once the score is out, the user will also get the recommendations for ancient scriptures and their related story which can help him/her heal fast. Routine affirmations based on the condition of user to generate positive thoughts. Also an intelligent bot or users on backend as strangers to chat with users feeling lonely can be a great feature.
So treatment part includes consultation part & self-treatment part. The consultation part is to give intelligent recommendations and then track the treatment progress. The self-treatment part focused on changing the mindset of user by changing the type of things they consume. Using spiritual learnings like that of Bhagwat Geeta.
- Privacy of user: Since, the data tracking on real-time and continuous monitoring may lead to privacy violation of users and various laws but if the tracking is not available, the precision and accuracy won’t be maintained. Results can be questioned.
- Data storage and handling: If implemented, the amount of data measured will be huge and hence a high-tech infrastructure and management system like that of giants like Google is must needed.
We can also say that Google already having a portion of user’s data, can come up with any product or service as they already have resources with them.
Every idea has a scope of improvement and hence I would appreciate your suggestions.