Inspiration

  • 1 doctor in Nigeria has to serve 10000 people, so doctors have up to 40 to 100 appointments daily.
  • From hospital to hospital, different digital data recording system makes it impossible to access patients past health records, financial and guarantor information before providing treatment slowly response time to even emergency treatments significantly, hence over 3000 people die daily due to emergencies.
  • Even worse, 70% of Nigerians depend on public healthcare systems where local healthcare workers still store their data manually, take diagnosis and then refer them to hospitals.

The theme of the Hackathon was "From Data to Prevention: AI as a health partner" But really? AI cannot be a health partner, where there is no data

Current AI Healthcare models are Western models trained on western people disease patterns, lifestyle and context. This would only lead to wrong diagnosis and further complications if we depend on this AI tools, this can be seen if a Nigerian asks an LLM questions about their health, answers are often going to be out of context medically.

Viewing these problems as one, we decided to build Nigeria's own Healthcare Intelligence System that

  • Solves the data problem at the grassroot through a simplified mobile interface for local healthcare workers
  • Is centralized and eases accessibility of this data from hospital to hospital, every hospital can access a patient records with a simple QR scan.
  • Leverages this collected data to build our very first custom healthcare AI model that uses the context of past medical records and local disease patterns to recommend simple preventive steps every patient can take daily to manage their health

What it does

With NHIS

  • Local Healthcare Workers uses their mobile phones to collect patients health data (Vitals, Symptoms, Guarantor information, financial information), AI then generates a summary report of the risk level of the patient, either high, medium or low, then recommends them to the next best step. Their information is then encrypted into a QR Code that is given to the patient.
  • At every and any hospital, with a simple scan of the QR Code, the hospital can see full information about a patient and through a Dashboard, a Doctor can see what case is an emergency and has high priority, hence knowing who to attend to first and faster. The ease of access to information and past records of every patient cuts down response time. Doctors then submits the outcome of their consultation or treatment of the patient on the platform and any Doctor anywhere who scans the QR Code can access this information as well as the patient's medical history.
  • On a mobile phone patients can also view their medical history and based on this, the AI model recommends simple lifestyle preventive steps to take daily and when to go to the hospital or clinic.

Over time by collecting and processing the healthcare data of millions of Nigerians, the AI model gets smarter in recommending preventive steps and recognising health patterns.

How we built it

Infra: We built our frontend (2 mobile and 1 desktop interface) on Typescript and built our Backend services using OOP with Java for scalability and speed. We integrated with GroQ AI(Llama) for the purpose of this Hackathon

Our process: Our build process was user focused, so we chose a simple mobile interface for Health workers, a desktop interface for Doctors and a mobile interface for patients.

We built incrementally, from a simple prototype of the frontend, to implementing the first few features (QR Scan, Data Collection), then to AI integration

Challenges we ran into

Skill issues, nothing more really

Accomplishments that we're proud of

  • We thankfully came 2nd
  • We learnt we had skill issues

What we learned

  • Leverage AI more in writing basic code so you can focus on logic and the real programming
  • Build more depth with both frontend and backend regardless of where you specialize so integration can be easy.
  • Don't be scared to rollback, it can just be a lifesaver
  • Solving a real problem and communicating it right always wins the pitch, always
  • Showing your demo even after a good pitch wins 1st place, don't just talk, show workings 🙃
  • Shout at your team members from time to time, it's good for building team spirit (lol)

What's next for Nigeria Health Intelligence System

We know what we have built as a mock project is crucial for Nigeria's development and will save millions of lives.

We hope we can find people with depth of experience in healthcare and data who will back us or take this up and ensure it becomes a real solution and Nigeria's foundational health infrastructure.

Beyond the build-pitch and win thing, we want to see this make real impact and hope that truly once this is built with the right partnerships the data collection starts at the grassroots and reaches the first 100,000 in the next year.

We hope that by next year at Cavista, the team will be proud to say that a project from last year is now transforming the industry and we hope Cavista can provide the support to make this happen.

Tech is transformational, we want to see the transformation this can birth for Nigeria as well.

Built With

Share this project:

Updates