OpenCare: Interoperable Hospital Ecosystem
This document outlines an open-source, role-based Flutter application designed for use in hospitals. The application includes AI-enabled vitals scanning, immutable clinical audit trails, and uses the FHIR standard for data interoperability.
>Overview:
OpenCare represents a mobile open-source ecosystem built on the principles of high-importance clinical environments where the integrity of data and the reliability of offline functionality are of utmost importance. Many hospital systems function as proprietary "black boxes" with little or no transparency; however, OpenCare implements the principles of Transparency, Interoperability, and Offline Reliability into its design process. OpenCare is built utilizing a Clean Architecture, utilizing GetIt and Provider technologies, to build distinct and secure dashboards for Nurses, Doctors, and Administrators.
>Key Innovations:
A) AI-Powered Vitals Scanner (OCR):
In order to reduce the friction associated with the manual data entry in the fast-paced environment of patient care, we integrated the Google ML Kit.
1.Functionality: Allowing nurses to use mobile phone cameras to point at traditional bedside monitors will result in the automatic extraction of heart rate, $SpO_2$, and blood pressure.
2.Impact: This will lead to the reduction of human error, as well as enable real time digital charting for all non-smart medical devices, resulting in up to an 80% increase in efficiency.
B) Immutable Clinical Audit Trail:
Transparency required by healthcare and open source solutions dictates that every change made to a patient's electronic medical record is cryptographically logged.
1. Accountability: Clinical staff can utilize a View History feature to see a historical audit trail when accessing any clinical record. The audit trail contains a timestamp with the 'Who', 'What', and the 'When' of the item being accessed.
2. Security: Provides data integrity and will prevent unauthorized or undocumented alteration to the patient's critical treatment plan(s).
C) FHIR Standardized Interoperability:
We have adopted HL7 FHIR (Fast Healthcare Interoperability Resources) standard for the export of our data.
1. Goal : This makes it possible for patient data to be shared with different global health systems rather than being kept "locked up" in one app.
2. Impact : The result is a fully interoperable healthcare network, which brings this project closer to having applied professional medical standardized processes.
>Technical Architecture
Uses clean architecture principles with a clear separation of concerns. Uses getIt for dependency injection and uses provider for reactive state management (reactive).
1. Frontend : The app is built using Flutter 3.10+, Material 3, and the flutter_animate package for a polished UX. The app utilizes the fl_chart package for displaying real-time medical visualizations.
2. Intelligence Layer : Google's machine learning kit is used for client-side OCR (optical character recognition), which allows for privacy-focused offline ingestions of data from medical monitors.
3. Reliability and Sync : The app is built using an offline-first approach, using SQLite for local persistence and Firestore for real-time cloud synchronization.
4. Data Standard : HL7 FHIR (Fast Healthcare Interoperability Resources) standard is used for data standardization, ensuring that the data is interoperable and globally ready for healthcare integration.
5. Performance Optimization : The app is optimized for performance by utilizing server-side indexing and capping the Firestore cache at 100MB ensuring the app remains responsive even on low-end hardware.