QnA Bot for Clinical Tabular Data

Abstract

Problem Statement

Medical jargon and terminology often means patients have difficulty in understanding and analysing the clinical data presented to them in their health reports. This means they can’t draw conclusions and make informed decisions about their care. Without the proper mitigation of certain illnesses, this leads to an increased risk in patient wellbeing.


Our Solution

Our solution is to build a questioning and answering bot for clinical tabular data. The patient will be able to upload their medicate report into a chatbot and begin asking questions immediately. The conversation between the user and the bot will mirror that between a patient and doctor, ensuring clear and concise phrases are used without technical jargon improving the user experience.


Achievement and Impact

With the completion of the project we had built a system that has solved our problem statement. Patients can now better understand their health records meaning they make more informed decisions about their care leading to better patient wellbeing. The improvement to patient well being not only improves the patient's quality of life but also will put less stress on the medical sector. By reducing the patient doctor communication time in mass can lead to innovation and medical discoveries as significant time is spent away from their patients and towards research. This ultimately will lead to a number of positive externalities for wider society.

Project Video

Team

Julia Goh

julia.goh.20@ucl.ac.uk

Researcher, Full-Stack Developer, Tester

Agnieszka Ostrowska

agnieszka.ostrowska.20@ucl.ac.uk

Researcher, Back-End Developer, Report Editor

Ben Schlagman

ben.schlagman.19@ucl.ac.uk

Researcher, Back-End Developer, Report Editor

Management

Our project timeline is illustrated in the Gantt Chart below, spanning over 2 academic terms.

Gantt Chart