While the discovery of new health information is always a good thing, it can lead to difficulties. Doctors and specialists are required to actively ensure they are informed about the latest innovations and news in their field. This includes the latest clinical trials, case studies, journal articles and research reports that may be relevant to their work. It’s impossible to keep up with this flood of information and, because of this, mistakes are sometimes made.

Jamal Zolhavarieh wants to make it easier for healthcare experts to make the right decisions by supplying them will all the necessary and important new information in a more user friendly and relevant way. Many clinicians already have access to a clinical decision support system (CDSS). These applications seek to link the observations of healthcare staff to a knowledge database to assist in determining accurate conclusions. These systems are helpful, but often have issues with the quality of information being displayed to the user.

The objective of Jamal’s project was to implement a ‘smart browser’ which would be able to find the highest quality knowledge and display it. The browser has access to a central knowledge base as well as electronic knowledge sources such as PubMed. In the image below, you can see the user interface with the ability to input keywords, as well as symptoms, and select from multiple filters to generate a list of precise articles.

The most important part of this project was developing a ranking system. Jamal used a text mining technique to assess how relevant the knowledge was in relation to the search query. This text mining method reads every word in each document that the search locates, and counts how many times words are repeated.

Each document is given two scores:

• Knowledge weight: A score based on how often the search terms are mentioned.
• Knowledge relevancy: A score based on common terms used in both the search query and document.

These two scores are then combined into an overall weighting, which determines the order in which they are displayed to the user. Those at the top of the list are considered more relevant to the user’s search terms. These metrics can be improved in the future to increase the accuracy and relevancy of search results. Jamal has also planned to interview practitioners and get their opinion on how relevant the articles are to their search terms, and if they agree with the weighting his system has assigned to the documents. Their feedback would help guide further development.

Jamal’s research fills an existing gap between existing CDSS software, and the end users. Ensuring that practitioners are getting up-to-date information is one thing, but being able to ensure the information is as relevant as possible is just as important. Relevant information means less time is wasted looking at unhelpful documents, increases the precision doctors have when making diagnoses and lowers the likelihood of mistakes being made. This project has the potential to make clinicians jobs that little bit easier, save them time, which in turn allows the healthcare system to operate more effectively.

Jamal is among a group of students who took part in the summer of research programme funded by Precision Driven Health. This month we are featuring a blog series examining these projects. While at an elementary stage and considered to be a ‘proof of concept’, these projects offer fresh insights into what the world of healthcare will look like when precision medicine is fully implemented.