Systematic literature reviews (SLRs) have become a major methodological tool in many areas of the health sciences and are particularly important in helping scientists and biopharmaceutical companies understand the current knowledge around a topic and identify future research and development directions. In the field of health economics and outcomes research (HEOR), researchers routinely conduct SLRs to understand the research landscape, synthesize evidence around unmet medical needs, compare the values of various treatment options, and prepare the design and execution of future real-world evidence studies.
Conducting an SLR involves synthesizing high-quality evidence from biomedical literature in a transparent and reproducible manner, seeks to include all available evidence on a given research question, and provides some assessment regarding the quality of the evidence. Currently, a SLR is often conducted manually, which is resource-consuming from both labor and financial perspectives. As biomedical literature grows at an unprecedented rate, advanced artificial intelligence solutions are an excellent approach to expedite SLR efforts.
Leveraging our strong expertise in biomedical literature mining, Melax Tech built an intelligent and comprehensive solution for SLR, which covers all necessary SLR steps, including study protocol setup (e.g., inclusion and exclusion criteria for literature search), literature retrieval, abstract screening, full-text screening, data element extraction from full-text articles, results export and visualization. Compared with existing SLR products, Melax Tech’s SLR solution is integrated with cutting-edge NLP algorithms to expedite time-consuming parts of SLR by providing machine learning-based recommendations.
For abstract screening, the NLP model makes recommendations on whether a particular article should be included for full-text review and provides supporting evidence (e.g., salient words occurred in inclusion and exclusion criteria and explanation models) for its prediction. For data element extraction, a hybrid pipeline consisting of deep learning algorithms and linguistic rules recognizes, extracts, and normalizes data elements from full-text articles in PDF format, including tables. Human experts validate the results and suggest changes if necessary. The NLP models are tailored to HEOR studies but can easily be expanded to other areas.
The system is currently being deployed to a major international pharmaceutical company.
To learn more about our SLR solution and how it implements NLP and machine learning technologies to make SLR efforts more efficient, accurate, and scalable across multiple therapeutic areas, request a demo today!