2018 Quro: Facilitating user symptom check using a personalised - - PowerPoint PPT Presentation
2018 Quro: Facilitating user symptom check using a personalised chatbot- oriented dialogue system Shameek Ghosh, Sammi Bhatia, Abhi Bhatia July 2018 Outline Understanding a patients journey Background of the Problem
Quro: Facilitating user symptom check using a personalised chatbot-
July 2018
Shameek Ghosh, Sammi Bhatia, Abhi Bhatia
Outline
Feel Sick Online Search Appointment Assessment Diagnosis
Patient jumps online, starts searching about symptoms, gets anxious from incorrect generic results. Patient books an appointment or walks in, describes the symptoms. Doctor asks about medical history, conducts examination. Doctor offers a diagnosis and explains context. Patient experiences symptoms, concern about certain conditions.
Prescription
Doctor prescribes medications, refers to specialist, requests tests
Driven By Doctor Driven By Patient Driven By Patient
Medius – Technology Nurturing Humanity
A Patient’s Journey TODAY
compliance-in-a-community-pharmacy-setting-Full-Final-Report-.pdf
Pre Consultation
84% of users of the internet incorrectly search for health
related information online1
During Consultation
75% of all GP visits are for minor ailments, repeat
prescriptions, referrals - a heavy burden on the healthcare system1.
Post Consultation
$1.2b per year is the cost to Australian healthcare
system due to non adherence of medication or treatment regimes - triggering readmissions2.
Problem
Leading causes
death
cancer 585,000
diagnostic error 251,000
heart disease 611,000
No of doctors per 1000 population 3rd Leading Cause of Death
0.59 0.64 1.1 2.15 3.31 3.59 3.9 4.78 India South Africa China United Kingdom United States of America Australia Germany Spain
Artificial Intelligence and developed by Medius Health.
bot for primary care.
likely causes of conditions and helps them decide w hat to do next and w here to go.
7M+ data points describing the relationships between…
Large-scale Data Extraction and Careful Curation
Salient Features of the Quro Chatbot and Triage System
Online instant medical triage: Online users want to know if their conditions require going to emergency care, visiting the GP, or remain at home and rest Improving engagement through an easy-to-use user interface and better user experience Using natural language processing to make sense of the user’s demands followed by sequential symptom question answering using a medical knowledge graph Ensemble models involving disease text embedding models for generation of a shareable pre-assessment report for a user for sharing with a GP
Overall architecture of Quro
Represent words in a continuous vector space For each word, the vector could reflect syntactic and semantic patterns such as the degree of similarity between words Neural Network is used to generate a vector matrix for word text in the corpus
Visualizing and clustering medical text
Word Embedding Models
Continuous evaluation process for condition pre-assessment: Current status
Evaluation Criteria At least 1 of the top 3 reported conditions is a correct assessment 2 out of 3 reported conditions were expected conditions by our in-house clinical experts Datasets used for testing 30 clinical vignettes curated by internal experts from primary case notes On-going evaluations across 10 diseases
Evaluation Results for triage pre-assessment: Current status
Initial evaluation using 30 patient vignettes in two different test criteria, showed an accurate outcome in 25 out of 30 cases (83.3%) and in 20 out of 30 cases (66.6%).
Investigated Disease Infectious Gastroenteristis (IG) Cholecystitis (CTS) Pelvic Inflammatory Disease (PID) Benign Prostatic Hyperplasia (BHP) Celiac Disease (CD) Ulcerative Colitis (UC) Menopause (MNP) Gastroesophageal Reflux Disease (GERD) Polycystic Ovarian Syndrome (PCOS). Irritable Bowel Syndrome (IBS) Urinary Tract Infectious (UTI)
On-going Study: List of diseases for word embeddings
Predict/True IG UTI IBS BPH GERD PCOS CTS PID CD UC MNP Precision Recall IG 12 1 1 0.667 0.857 UTI 9 1 1 1 2 0.818 0.643 IBS 4 10 0.833 0.714 BPH 1 12 1 0.75 0.857 GERD 13 1 0.928 0.929 PCOS 1 10 1 1 1 0.909 0.714 CTS 1 12 1 0.706 0.857 PID 1 11 1 0.846 0.846 CD 1 1 10 2 0.833 0.714 UC 1 1 1 1 10 0.625 0.714 MNP 2 12 0.923 0.857 Accuracy 0.791
Initial Multi-class prediction results using Embedding Models
Future Work: Planned immediate improvements
Scale context embedding model to multi-hundred-disease prediction for 150 diseases activated in the Quro system Integration of Quro knowledge graph with context embedding models Further evaluations using expanded set of clinical vignettes across 150 diseases
Takeaways
Consumer focussed engagement for understanding user symptoms, their needs, and provide valuable information to physicians for further inquiry Proactive dynamic collection of illness narrative over time, prior to doctor appointments Use of NLP entity recognition and relation extraction algorithms to determine initial entry points in knowledge graph Graph reasoning engine for optimal sequential question answering Automated data collection and expert driven primary case collection for development of disease prediction models
Thank you
Head Office Level 25, Tower Three, International Towers Sydney, Barangaroo https://mediushealth.org/ Phone +61 430 450 204 shameek.ghosh@mediushealth.org