2Daybetes

Imagen del trabajo dentro de matlab en donde se muestra de manera general las funciones que se utilizaron.
Petición de historial clínico dentro de nuestro cuestionario
Petición de datos generales por parte de nuestro cuestionario
Primera imagen del test en donde se especifican la privacidad de los datos obtenidos
Información general de la diabetes en mexico
Primer caso: Tienes una baja posibilidad de tener diabetes
Segundo caso: Tienes una alta posibilidad de tener diabetes

Inspiration

This web project was born from personal experiences with family members facing health issues related to diabetes, a condition that could have been prevented with early assessments. These situations alerted us to the crucial importance of timely detection of the risk of chronic diseases like diabetes, which, according to data from the International Diabetes Federation, affects 12.8% of Mexican adults. However, undergoing comprehensive medical evaluations regularly remains a challenge for much of the population. It is estimated that around 70-80% of diabetes cases in Mexico are undiagnosed.

What it does

Our program, 2Daybetes, is an innovative web tool designed to assess diabetes risk in a simple and accessible manner. Through an intuitive questionnaire, the system collects user data such as age, weight, medical history, and blood glucose levels. Using machine learning algorithms, the system analyzes this data and provides a prediction of the risk of developing diabetes. Additionally, it offers informational resources and personalized recommendations for the prevention and management of the disease.

How we built it

We built 2Daybetes using various advanced technologies:

MATLAB Net Fitting: We implemented a regression neural network using MATLAB, trained with a Kaggle dataset to make accurate predictions about diabetes risk. Figma: We used Figma to design interactive mockups that helped us create an intuitive and collaborative user interface. Python http.server Module: We deployed the webpage locally using this module, allowing us to conduct iterative testing and development without the complications of complex web server configurations.

Challenges we ran into

During the development of 2Daybetes, we faced several challenges:

Training the machine learning model: Implementing and optimizing the neural network in MATLAB was a complex process that required multiple iterations and adjustments. User interface design: Creating a user experience that was both intuitive and informative involved numerous design changes and tests. Technology integration: Ensuring that the different technologies (MATLAB, Figma, and Python) worked cohesively was a significant technical challenge. Accomplishments that we're proud of

We are proud of several achievements in the development of 2Daybetes:

Development of an effective neural network: We successfully trained a neural network capable of making accurate predictions about diabetes risk using real data. Design of an accessible interface: We created an easy-to-use web platform that provides crucial information and personalized recommendations for users. Implementation of an integrated solution: We integrated multiple technologies to create a cohesive and functional tool that can be used by anyone with internet access.

What we learned

Throughout the project, we learned a lot about:

Machine learning and its application in healthcare: We gained experience in implementing machine learning algorithms and their ability to process and analyze health data. User-centered design: We understood the importance of creating interfaces that are accessible and useful for users, especially in the healthcare context. Interdisciplinary collaboration: We learned to work in a multidisciplinary team, integrating knowledge from engineering, design, and medicine to develop an effective solution.

What's next for 2Daybetes

The future of 2Daybetes includes several plans for expansion and improvement:

Expansion of the dataset: We plan to incorporate more data to enhance the accuracy and robustness of our predictions. Additional features: We aim to add new functionalities, such as continuous health tracking and personalized recommendations based on the user's history. Collaborations with healthcare professionals: We seek to establish partnerships with doctors and health organizations to validate and further improve our tool. Expansion to mobile devices: We plan to develop a mobile application to make our tool even more accessible.