CLEAR YOU MIND — bridge the time till your psychotherapy consultation

8 min readJun 27, 2023

This project was carried out as part of the TechLabs “Digital Shaper Program” in Dortmund (winter term 2022/23).

In a nutshell:

“Clear Your Mind” is a platform designed to offer urgent and quick psychological support to individuals who are struggling with mental health issues.

Traditional methods of accessing therapy often involve long wait times, leaving many individuals feeling helpless and stuck.

Our platform offers a simple and effective approach to help users gain a deeper understanding of their emotions and inner world, using methods such as journaling and mood tracking. Through these techniques, users can gain valuable insights into their thought patterns and emotions, and take positive steps towards improving their mental well-being in a short period of time.

Introduction:

Our platform is not only about supporting people with mental health issues. The main problem we strive to address is social inequality. Why do wealthier people have the security of a timely therapy place, although mental illness is an issue in all social classes? We won’t be able to change the health system any time soon, so we need to find other ways. The majority of our population owns an internet-enabled smartphone. So why not offer simplified therapies via the internet?

First of all, our project is meant to address the long wait times patients experience before their first consultation with a psychotherapist. We have designed our platform to be compatible with all clinical pictures and easy to use, providing patients with simple tools to take care of their mental health and better prepare them for their upcoming consultation. Our web application features a comprehensive questionnaire to assess depression severity, a content page with materials to aid in treatment, a forum for peer support and information exchange, and a mood tracker and diary function for patients to monitor their emotions and reflect on their experiences.

With these tools at their disposal, patients can make the most of their bridging time by taking proactive steps towards improved mental health. By using the mood tracker and diary function, patients can identify patterns and triggers that may contribute to negative moods or behaviors, empowering them to make meaningful changes in their daily lives. Ultimately, our platform seeks to provide patients with the support they need to take control of their mental health and achieve long-term well-being.

Methodology:

The initial scope of our project was quite large, which required us to prioritize our tasks. To achieve this, we used the MoSCoW prioritization technique, which helped us divide our features into four categories based on their importance. We then turned these into product backlog items (PBIs) to better organize and manage our tasks. To make the PBIs as clear as possible to everyone, one of our Data Scientists Okan, with the help of our product owner Yousra, wrote the requirements for them.

To improve our workflow, we implemented a simplified Scrum routine. Scrum is an agile project management framework typically used in software development but can be adapted to different projects. We utilized ClickUp to organize our PBIs on a Kanban board and divide them into sprints.

To track our progress and address any issues, we held weekly meetings. During mid-sprint, we conducted basic Scrum standup meetings to update each other on our progress since the last meeting and discuss our plans until the end of the sprint. Everyone also got the opportunity to share any struggles or roadblocks he is dealing with at the moment. At the end of each sprint, we held a Scrum review meeting to present our completed tasks, followed by a retrospective meeting to discuss any necessary adjustments to our workflow. Finally, we held a sprint planning meeting to decide on the next set of PBIs to tackle.”

To streamline our efforts Mariusz, who has some professional experience in software development, created a few documents on click-up that included guides on writing requirements, user stories and setting up VS Code and GitHub.

Development Process:

Our journey in developing our web application has been full of challenges. One of the initial hurdles we faced was the lack of a UX Designer on our team to help us create some mock-ups for the front end. However, we were determined to make our application visually appealing and decided to take matters into our own hands. Our web developer,

Mariusz, took the initiative to learn Figma through a short tutorial and created the mock-ups himself.

For the actual implementation of the mock-ups, we chose React as our development framework. While working with React, we encountered several compilation errors that proved to be quite daunting. Nevertheless, with our team’s mutual support and perseverance, we managed to overcome these obstacles and continue making progress towards our goal.

After looking through different mental health questionnaires available online, Yousef decided to pick the most suitable ones which were simple enough for both the users and the task of developing it. After selecting the Beck Depression Inventory (Beck et al., 1961), he developed the frontend page using React. The data for the questionnaire were stored in a local .js file temporarily, since the backend was still on the backlog.

The biggest challenge in incorporating the concepts learned in the data science program into a practical implementation was identifying a suitable use case for their application. However, after careful consideration, the most promising approach in our opinion was the development of a recommender system. The proposed system involves utilizing the responses of existing users who have previously completed a questionnaire and shown an interest in specific content available on Clear Your Mind. Subsequently, when a new or existing user completes the questionnaire for the first time, their responses are compared with those of the existing users to generate personalized recommendations.

Using cosine similarity, the recommender system can build a matrix and give a similarity rating between each user based on their answers to the questionnaire. With this matrix available, the system can easily see who the topmost similar users are, fetch the content they liked, and recommend it to the new user.

Data science was also used by Okan for the implementation of the mood tracker. After Okan’s implementation, Yousef implemented a stacked bar graph to visualize the mood tracker data of a user. The stacked bar graph presents the users’ mood tracker entries,

categorized according to the day of the week. This enables the users to identify any possible patterns or trends in their mood over time. For example, some people might find that their moods are usually worse on Mondays, or better on Saturdays. By analyzing the trends and patterns in their mood through the stacked bar graph, users can potentially identify specific triggers in their daily life that may affect their mood. This can help them develop better coping mechanisms and strategies to manage their emotions effectively.

In addition to identifying triggers, the mood tracker data can also be applied to track changes in mood over time. This feature enables users to visualize their progress and determine if their mood is improving or deteriorating. If the data suggests a decline in mood, users can seek immediate help. Conversely, if the data indicates an improvement in mood, it may serve as an indicator that their therapy or strategies are working effectively.

Results:

In the end of the project phase, we didn’t achieve everything we initially planned but are still satisfied with the current version of our web application.

We managed to implement two of our front-end features the questionnaire and the forum. The forum has an integrated backend that allows us to save posts in our MySQL database and retrieve them back to fetch them on the user interface. Other front end features lack the back end for now but we are going to utilize the models provided by Yousra and integrate the backend to them shortly.

For data science, we were able to create a recommender algorithm that uses cosine similarity to find similar users based on their questionnaire answers (dummy data). The final result is to provide a list of 10 users with the highest similarity based on this dummy data. Additionally, we use mood tracker data (dummy data) and plot graphs that show mood tracker data in a bar chart by day of the week, and a line graph that shows mood changes over the last 30 days.

In future iterations of the project, the recommender algorithm will be enhanced by fetching real data from the backend, which will enable it to calculate similarities and return a list of 10 users. This will greatly improve the accuracy of the algorithm by allowing us to include multiple questionnaire data as input. Additionally, we are planning to implement authentication, deploy our website, and finalize more front-end features. For example, we plan to extend our forum by allowing users to post comments, which will further enhance our platform’s engagement and interactivity.

With these improvements, we believe our project will be able to provide even more personalized and tailored recommendations to our users.

Github repository:

https://github.com/TechLabs-Dortmund/clear-your-mind

Team:

Web Development: Mariusz, Seget [mariuszseget77@gmail.com, https://www.linkedin.com/in/mariusz-seget/?locale=en_US, Scrum Master, Full Stack Developer, UX Designer ]

Data Science: Yousef, Hamadah [yhamadah@wwu.de, Data Scientist, Front End Developer]

Data Science:Okan, Mutlu [okan.mutlu@tu-dortmund.de, https://www.linkedin.com/in/okan-mutlu-8a2a7b252/ Data Scientist]

Web Development: Yousra, Grad [yousra.grad@tu-dortmund.de, https://www.linkedin.com/in/yousra-grad-036274211/, Backend Developer, Product Owner]

Mentor:

Nils Jannasch, Web Development Mentor

Timo Behrendt, Project Manager

Tom Stein, Web Development Mentor

Figures & Images: