AortaAid

Inspiration

The aorta, the largest artery in the body, plays a crucial role in circulating oxygen-rich blood from the heart to the rest of the body. Aortic diseases, such as aneurysms and dissections, can be life-threatening and often require prompt diagnosis and treatment. Given the high stakes of cardiovascular diseases and the pivotal role that an effective diagnosis plays in patient care, we aimed to create a tool that empowers medical caregivers with advanced diagnostic capabilities, ultimately improving patient outcomes. We were motivated by the potential to leverage AI technology to bridge gaps in healthcare, especially in under-resourced areas where specialist doctors might not always be available. By providing caregivers with this accessible diagnostic tool, we hope to reduce the burden on healthcare systems and improve patient care efficiency.

What it does

AortaAid is a web application designed to assist medical caregivers in diagnosing patients with cardiovascular issues, furthermore determining the “risk factor” of each patient. By inputting patient data into the application, nurses can receive an analysis generated by an AI trained on numerous cardiovascular datasets present on Kaggle. This data-driven method of analyzation, combined with the medical expertise of doctors and nurses, can lead to timely diagnoses that could ensure appropriate medical interventions. The application offers a user-friendly interface where nurses can enter various patient metrics such as BMI, age, and smoking status, and furthermore enter cardiovascular issues such as heart disease and risk of stroke. Once all of the metrics are given, the program’s trained AI calculates the most impactful factor in the case of a negative diagnosis, and provides advice to better the patient’s cardiovascular health.

How we built it

We split our app into two main sections: The frontend app and our AI models (backend). For the frontend app, we used a Next.JS framework to use React.js, TailwindCSS, and Typescript to build out the user interface. For the backend application of our project, we used Python and Scikit to train five AI models to help predict and analyze risks in various cardiovascular diseases. Then, we connected the two using FlaskAPI and Axios, creating REST requests to the backend from the frontend by transferring JSON data.

Challenges we ran into

One major hurdle was ensuring the quality and consistency of the data from Kaggle, which required extensive preprocessing and cleaning. We also faced challenges related to ensuring the security and privacy of patient data.

Accomplishments that we're proud of

We are proud of successfully integrating a sophisticated AI model into a practical tool that can be used in real-world clinical settings. We are also proud of the collaborative effort that went into this project, and the learning experience that came for all three of us.

What we learned

Through this program, we gained valuable insights into the complexities of medical data and the importance of data quality in training effective AI models. We also learned a great deal about the challenges of developing effective healthcare applications.

What's next for AortaAid

In order to get more precise results, we intend to collect data from additional hospitals in the future. We hope to gain access to a wide variety of patient data by working with more healthcare organizations, which will improve the accuracy and resilience of our AI models. Our ability to fine-tune our diagnostic algorithms and make sure they work well for a variety of populations and clinical contexts will be enhanced by the growth of data sources. Furthermore, we plan to broaden our scope beyond cardiovascular health in order to create a comprehensive tool for medical caregivers. Incorporating diagnostics for a broader range of illnesses will enable nurses and other healthcare professionals to diagnose patients more rapidly by employing more precise measurements. AortaAid has the potential to become a comprehensive system that supports multiple facets of patient care, ranging from continuous monitoring to early detection, by incorporating extra health indicators and diagnostic capabilities.

Built With

• kaggle
• next.js
• python
• react
• scikit-learn
• typescript