AI Powered Decision Support System for Assisting Medical Staff
About the Client
A client is a personalized medicine service provider, aimed at making cutting edge science and advanced clinical experience available for medical practitioners.
Amid the COVID-19 pandemic, the global healthcare system is facing new challenges. A shortage of medical staff and a high workload of hospitals slow down the provision of care to patients, which makes it ineffective to combat the virus spreading.
The client wanted to create a decision support system that would help doctors examine patients, record health data into EMR and get recommendations on treatment for COVID-19 cases.
We developed the decision support system to help medical staff of various qualification levels who are involved in the fight against COVID-19 to prescribe and monitor treatment processes faster and efficiently. The doctor uses DSS to record medical information collected according to the examination protocol for a patient suspected of having COVID-19. After filling in all the data, the system automatically synchronizes the information from the DSS to the hospital’s medical information system. The system allows doctors to quickly update the treatment protocol, forcing medical staff to follow the protocol and receiving recommendations for the treatment of patients with COVID-19.
The developed DSS allows addition and configuration of medical protocols and implements doctor AWP (any willing provider). The protocol consists of rules defined by experts that allow recording the examination results as a patient’s EMR and suggests the next steps upon a patient’s treatment history. The first protocol implemented during pilot training was the COVID-19 diagnosis treatment protocol that a doctor must follow during the physical examination.
The DSS system also provides recommendations that doctors can apply to treat a sick patient. It refers to the protocol, assesses the patient’s status, and provides the doctor with recommendations when hospitalization is required or when the patient can continue treatment at home.
For the quick access to the AWP, we created a web application that can run from any device connected to the Internet. It synchronizes with the hospital’s MIS via API and allows rapid retrieval of patient data. The doctor AWP UI is generated upon configured rules in the protocol.
After the pilot was integrated and configured, we started rolling out the project to other hospitals as per client’s requirements.
Special attention has been paid to the security subsystem which allows configuring roles, granularly give permissions even on a record level, and logs all actions in the system.
Let's discuss your idea!
The DSS is implemented on top of a custom rule-based engine, which is mainly written in Python. PostgreSQL was used as a data storage. For dynamic UI – React.