What is Oscar?
This project aims to create Oscar (Oncology Search Clinician Assistant Registry), a search engine for finding relevant indicative treatments for a specific child patient based on the genomic profile of their cancer.
The information being searched is:
- ~300,000 clinical trials from ClinicalTrials.gov; and
- ~25 million medical articles from PubMed.
The specialised search capabilities of Oscar mean clinicians can find the right treatments for their patients with higher accuracy and less time/effort, ultimately leading to better patient outcomes for the most vulnerable patients - kids with cancer.
What does Oscar do?
Oscar is capable of ingesting and tracking all medical publications and clinical trials globally. From these vast amount of data and using advanced AI algorithms, Oscar is able to extract gene, drug and cancer type information. Based on this information, Oscar locates relevant treatments tailored to individual patients. Oscar also provides tools for the effortless exploration of the treatment landscape. This allows doctors to be provided with information on the best treatments for their patients by an automated system.
Who made Oscar?
- The Australian e-Health Research Centre at CSIRO created Oscar.
- Funding was provided by the Queensland Hospital Foundations.
- Clinical expertise provided by Queensland Children's Hospital.
Team
A/Prof Bevan Koopman
Project lead
Dr Shengyao Zhuang
AI Researcher - Ranking
Dr Hoa Ngo
AI Researcher - Natural Language Processing
Simon Gibson
Software Engineering
Tracey Wright
Software Engineering
Prof Guido Zuccon
Academic Partner
Oscar in Action
Oscar main search page.
Oscar Knowledge Graph visualisation to help understand associations between genes, drugs and cancers.
Relevant Links
- Video presentation from the SIGIR search engine conference titled Precision medicine search for paediatric oncology.
- Research paper. B. Koopman, T. Wright, N. Omer, V. McCabe, and G. Zuccon. Precision medicine search for paediatric oncology. In SIGIR, 2021.