BioASQ Synergy: a dialogue between question-answering systems and biomedical experts for promoting COVID-19 research

Anastasia Krithara ¹, Anastasios Nentidis ^1,2, Eirini Vandorou ¹

¹Institute of Informatics and Telecommunications, National Center for Scientific Research "Demokritos", Athens 15341, Greece.
²Department of Informatics, Aristotle University of Thessaloniki, Thessaloniki 54124, Greece.
³Institute of Biosciences and Applications, National Center for Scientific Research "Demokritos", Athens 15341, Greece.
⁴Department of Biology "Charles Darwin", Sapienza University of Rome, Rome 00185, Italy.
⁵Lithuanian University of Health Sciences, Kaunas 44307, Lithuania.
⁶Columbia University Vagelos College of Physicians and Surgeons, New York, NY 10033, United States.
⁷Barcelona Supercomputing Center, Barcelona 08034, Spain.
⁸NHS Improvement, London SE1 8UG, United Kingdom.
⁹AstraZeneca Pharmaceuticals, Wilmington, DE 19803, United States.
¹⁰Ochre Bio Ltd, Oxford OX4 4GB, United Kingdom.
¹¹Hospital Vithas Madrid la Milagrosa, Madrid 28010, Spain.
¹²St George's University of London, London SW17 0RE, United Kingdom.
¹³Institute for Bioinnovation, BSRC "Alexander Fleming", Vari 16672, Greece.
¹⁴Clinical Genomics and Pharmacogenomics Unit, 4th Department of Internal Medicine, Attikon Hospital, Medical School, National and Kapodistrian University of Athens, Athens 12462, Greece.
¹⁵Molecular Biology Division, Biomedical Research Foundation of the Academy of Athens, Athens 11527, Greece.
¹⁶Department of Informatics and Telecommunications, National and Kapodistrian University of Athens, Athens 16122, Greece.

⁰Institute of Informatics and Telecommunications, National Center for Scientific Research "Demokritos", Athens 15341, Greece.

Journal of the American Medical Informatics Association : Jamia +

|

August 24, 2024

View abstract on PubMed

Summary

The BioASQ Synergy research process enhances expert-AI interaction for biomedical question answering. This iterative approach aids researchers in discovering solutions and navigating scientific literature effectively.

Area Of Science

Biomedical Informatics
Artificial Intelligence
Knowledge Discovery

Background

Biomedical research generates vast amounts of data, making it challenging for experts to stay updated.
Automated question-answering (QA) systems offer potential but require expert validation and refinement.
Bridging the gap between human expertise and AI capabilities is crucial for scientific advancement.

Purpose Of The Study

To introduce the BioASQ Synergy research process, fostering collaboration between biomedical experts and automated QA systems.
To enable systems to answer emerging questions, with expert feedback driving iterative improvement.
To facilitate incremental understanding and solution discovery in complex biomedical problems.

Main Methods

A novel iterative research process involving biomedical experts and QA systems.
Systems provide answers to expert-generated questions.
Expert assessments and new questions are fed back to the systems for continuous learning.

Main Results

The BioASQ Synergy approach assists researchers in navigating scientific resources.
Experts reported high satisfaction with the quality of system-generated answers.
The study indicates the utility of such systems in addressing open research questions.

Conclusions

BioASQ Synergy aims to provide experts with personalized access to the latest findings in rapidly growing biomedical literature.
The system facilitates a continuous dialogue between experts and AI to address complex research questions.
An initial proof-of-concept demonstrated the usefulness of the approach for both experts and systems.

Keywords:

biomedical experts biomedical information question answering

Related Concept Videos

Enzyme-Linked Immunosorbent Assay

In 1971, Peter Perlman and Eva Engvall developed an Enzyme-linked immunosorbent assay (ELISA or EIA). ELISA differs from western blot in that the assays are conducted in microtiter plates or in vivo rather than on an absorbent membrane.
There are many different types of ELISAs, but they all involve an antibody molecule whose constant region binds an enzyme, leaving the variable region free to bind its specific antigen. Enzyme-substrate reaction allows the antigen to be visualized or...