CareGuide

Enhancing home healthcare with AI. Our app bridges the communication gap between doctors, nurses, and patients by providing a 24/7 at-home nurse with access to real-time support and personalized care.

Comment

Inspiration

CareGuide was developed by our team in response to the pressing problem of medical errors in healthcare. A Johns Hopkins study found that medical errors account for over 250,000 deaths each year in the US, making them the third most common cause of mortality. Incomplete patient information and poor communication are frequent causes of these mistakes. Furthermore, 42% of patients experience anxiety when it comes to taking care of their health at home, underscoring the need for improved support networks. Our goal was to develop a solution that would lower these risks and enhance patient outcomes by streamlining communication and offering thorough health overviews.

What it does

CareGuide is an AI-powered mobile app that enhances home healthcare by providing post diagnosis patient care. We help patients based on the interactions between patients, nurses, and doctors while also integrating the information from the discharge notes and Electronic Health Records (EHRs) of the patients. We are tailoring personalized care plans to individual patient needs based on their data and AI models which were specifically trained for this job. This combination of features ensures that patients receive timely, accurate, and personalized care, ultimately improving their overall health experience.

How we built it

Our application development process employed React.js for a user-friendly front-end interface, complemented by extensive CSS styling. The back-end infrastructure was constructed using Python with Flask serving as the server framework. We incorporated advanced speech-to-speech technology and utilized Retrieval-Augmented Generation (RAG) to enable Large Language Models (LLMs) to access and interpret patients' clinical notes and doctor-patient conversations. Our cloud infrastructure, secured with AWS Client-Side Encryption, ensures the secure storage and processing of data, thereby upholding stringent data privacy standards and complying with HIPAA healthcare regulations.

Challenges we ran into

During the hackathon, we encountered significant technical challenges, particularly with setting up local AI using Intel AI on the provided hardware to create a medical specific model. Ensuring patient data privacy through local AI is critical, and regulatory compliance demands rigorous testing and implementation of robust security measures, which is difficult to achieve under tight deadlines. Consequently, we pivoted to alternative technologies to ensure functionality. Prioritizing features that address the actual needs of healthcare professionals was also challenging. However, we successfully conducted interviews with five doctors, gathering valuable feedback that helped us refine our product to better meet their needs.

Accomplishments that we're proud of

We are proud of the positive feedback received during the testing phase, which validated our approach and highlighted the potential impact of CareGuide. Successfully improving care received by patients with anyone around them and it allowed family members to focus more on patient health. Furthermore, providing real-time support and personalized care significantly improved patient satisfaction, demonstrating the effectiveness of our solution. We also cold called doctors and got validation from 5 of them for the product and even had positive feedback from fellow hackers.

What we learned

Participating in this hackathon taught us the importance of user-centered design, as engaging with end-users throughout the development process was crucial in creating a practical and effective solution. We also learned the value of building a scalable app that can grow with the increasing demand for home healthcare. Additionally, the experience highlighted the importance of effective team collaboration and incorporating diverse feedback to enhance our project.

What's next for CareGuide

Looking ahead, we plan to expand CareGuide’s features by adding very strict data privacy laws to avoid any HIPPA compliant issues. We also want to improve the model efficiency for medical information and keep training it on further datasets. We will also start monitoring of vital signs to help patients avoid any upcoming health complication. We aim to scale up the deployment to more regions and healthcare systems to reach a broader audience. Continuous improvement based on user feedback and technological advancements will ensure that CareGuide remains at the forefront of home healthcare innovation, providing the best possible care for patients.

References

  1. Centers for Disease Control and Prevention, National Center for Health Statistics. About Multiple Cause of Death, 1999–2020. CDC WONDER Online Database website. Atlanta, GA: Centers for Disease Control and Prevention; 2022. Accessed February 21, 2022.
  2. National Safety Council analysis of Centers for Disease Control and Prevention, National Center for Health Statistics. National Vital Statistics System, Mortality 2018-2022 on CDC WONDER Online Database, released in 2024. Data are from the Multiple Cause of Death Files, 2018-2022, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Accessed at http://wonder.cdc.gov/mcd-icd10-expanded.html
  3. Brady RE, Braz AN. Challenging Interactions Between Patients With Severe Health Anxiety and the Healthcare System: A Qualitative Investigation. J Prim Care Community Health. 2023 Jan-Dec;14:21501319231214876. doi: 10.1177/21501319231214876. PMID: 38041442; PMCID: PMC10693786.

Built With

Share this project:

Updates