Mudita

Mudita

Take the first step towards a healthier mind and happier you

Mudita

Mudita

Take the first step towards a healthier mind and happier you

The problem Mudita solves

Unlike physical illness, which has visible symptoms, mental illness is difficult to detect in the early stage and hence is often ignored. According to Lancet, more than 1 Billion people live with mental illness. An important point to draw attention to is that half of all mental illnesses begin to show symptoms by age 14. However, due to invisibility and lack of awareness, these symptoms are not detected in their early stage, which leads to serious mental illness later.

We aim to aid in the detection of these symptoms in their earliest stage possible so that serious illness can be prevented. To achieve the same, we conduct a survey of a user and predict what illness symptoms the user is possibly suffering from (if any) using our ML model. Then, based on the result, we recommend doctors that they can consult with.

One more challenge that we want to address is that people do not understand what exactly anxiety, depression, OCD, etc., is, due to which they never know that they are suffering from those illnesses. Thus, our survey is designed such that a single survey can make predictions for more than one illness.

Mudita can be very effective if schools, colleges, employers, etc., use this platform to conduct surveys of students and employees on regular basis.

Challenges we ran into

Problems we ran into:

  1. Finding relevant questions to ask in the survey. We studied a lot of articles and selected some best questions that we found appropriate.

  2. Finding a proper dataset to train our model. The way we are addressing mental health issues is unique, so we could not find any existing dataset for our purpose. However, we generated a dummy dataset to train our model.

  3. Initially, we were asking the same set of questions each time to each user. But then we implemented the survey such that the next question depends on the response to the current question. This way, the survey is more relevant to the user.

We are sure that if we get the data we require from authentic sources, our solution will work gracefully in real life.

Discussion