2019 June JAMIA Journal Club Webinar

June 13, 2019
Free for AMIA members; $50 for non-members.
Piotr Przybyla, PhD

Quantifying risk factors in medical reports with a context-aware linear model

Lead author Piotr Przybyla will discuss this month's JAMIA Journal Club selection:

Przybyła P, Brockmeier AJ, Ananiadou S. Quantifying risk factors in medical reports with a context-aware linear model. J Am Med Inform Assoc. 2019 Mar 6. pii: ocz004. doi: 10.1093/jamia/ocz004. [Epub ahead of print]  [Article]


Piotr Przybyła, PhD
Assistant Professor
Institute of Computer Science
Polish Academy of Sciences
Warsaw, Poland

Dr. Piotr Przybyła is an assistant professor at the Linguistic Engineering Group at the Institute of Computer Science, Polish Academy of Sciences (ICS PAS). He holds a PhD in Computer Science from ICS PAS and specializes in developing methods of Natural Language Processing (NLP) and Machine Learning (ML) for real-world applications. During his stay as a research fellow at the National Centre for Text Mining of University of Manchester, he focused on the domain of biomedical text, publishing NLP and ML solutions for the challenges of literature search, systematic reviews, precision medicine and information extraction from EHRs. Currently at ICS PAS, Dr Przybyła is leading a project on automatic credibility assessment for online media (HOMADOS).


  • 35-minute discussion between the authors and the JAMIA Student Editorial Board moderators including salient features of the published study and its potential impact on practice.
  • 25-minute discussion of questions submitted by listeners via the webinar tools and moderated by JAMIA Student Editorial Board members


JAMIA Journal Club managers and monthly moderators are JAMIA Student Editorial Board members:

Kelson Zawack, PhD, Postdoctoral Fellow, Biostatistics Department, Yale University, New Haven, CT

Tiffany J. Callahan, MPH, PhD Candidate, Computational Bioscience Program at the University of Colorado Denver Anschutz Medical Campus, Aurora, CO


Daniel Feller, MS, PhD Candidate in Biomedical Informatics, Columbia University, New York, NY


The PubMed citation for the paper under discussion is:

Przybyła P, Brockmeier AJ, Ananiadou S. Quantifying risk factors in medical reports with a context-aware linear model. J Am Med Inform Assoc. 2019 Mar 6. pii: ocz004. doi: 10.1093/jamia/ocz004. [Epub ahead of print]  [Article]

Student Access

Students who are not AMIA members, but whose academic institutions are members of the Academic Forum, are eligible for a complimentary JAMIA Journal Club registration. Please contact Susanne Arnold at susanne@amia.org for the discount code. In the email, please include: full name, Academic Department, and the primary Academic Forum representative of that Academic Department. Note that AMIA Student memberships are $50, which allow access to JAMIA, all JAMIA Journal Clubs, and other webinars of interest to the biomedical informatics community. 

Statement of Purpose

It has been shown in several studies that free text in electronic health records frequently contains valuable information that is not available in structured elements (codes) but could be extracted using Natural Language Processing techniques. Unfortunately, a mere occurrence of a name of a medical concept (eg. a certain disease) in a text does not clearly define its relation to a patient. One needs to take into account the context of this occurrence, including expressions of negation, certainty, severity, numerical quantities, etc. Given that the contribution of these contextual factors differs between medical concepts they refer to, previous solutions were very focused, dealing only with the aspects that are commonly affecting the condition of interest.

In the presented study, we sought to automatically quantify the mortality risk associated with all risk factors mentioned in an EHR text. Given the multitude of relevant concepts in such a task, the issue of taking into account the influence of the context for each of them was particularly challenging. We formulated it as a multi-task learning problem and solved through linear regression with appropriately designed regularization. The evaluation showed the proposed solution effectively assessing risk even for previously unseen factors and the overall performance of risk quantification reaching the level of human annotators' agreement. The obtained results point to the great potential of linear regularized methods in highly multi-dimensional and multi-task challenges, such as information extraction from electronic health records.

Target Audience

The target audience for this activity is professionals and students interested in biomedical and health informatics.

Learning Objectives

The general learning objective for all of the JAMIA Journal Club webinars is that participants will

  • Use a critical appraisal process to assess article validity and to gauge article findings' relevance to practice

After this live activity, the participant should be better able to:

  • Describe the extraction of risk factors from electronic health records as a machine learning task
  • Identify how the linguistic context of a medical concept mention can affect its importance
  • Explain the role of a regularization term in the design of a linear model

This JAMIA Journal Club does not offer continuing education credit.

In our dedication to providing unbiased education even when no CE credit is associated with it, we provide planners’ and presenters’ disclosure of relevant financial relationships with commercial interests that has the potential to introduce bias in the presentation: 

Disclosures for this Activity

These faculty, planners, and staff who are in a position to control the content of this activity disclose that they and their life partners have no relevant financial relationships with commercial interests: 

JAMIA Journal Club Faculty: Piotr Przybyla
JAMIA Journal Club planners: Michael Chiang, Kelson Zawack, Tiffany J. Callahan, Daniel Feller
AMIA staff: Susanne Arnold, Pesha Rubinstein

Contact Info

For questions about webinar access, email Susanne@amia.org.