A research milestone for World Heart Day

A contribution carried forward

Celebrating scholarship with wider reach

Elsevier has included this article in its World Heart Day collection, publicly listing it as the contribution from American Heart Journal Plus: Cardiology Research and Practice. The collection brings together selected cardiovascular articles and invites readers to explore and share them. Its featured access period runs through October 31, 2026; this paper is also published as an open access article.

We celebrate Dr. Sherry-Ann Brown, the paper's first and corresponding author, together with her coauthors and the CAIPE Research Team Investigators. This publisher feature creates a welcome occasion to recognize their work, introduce the research to additional readers, and invite thoughtful discussion about its possibilities and limitations.

The public collection is named World Heart Day. Our celebration marks that documented feature and the opportunity it offers for research education. The most meaningful response is to read closely, give the full team credit, and ask how the ideas might inform the next carefully designed investigation.

The published study

From imaging reports to research-ready measurements

This retrospective feasibility study examined rule-based natural language processing in Python and R across more than 26,000 imaging reports from 4,626 cancer survivors, spanning 2000–2020. The five modalities were transthoracic echocardiography (TTE), transesophageal echocardiography (TEE), multigated acquisition (MUGA), single-photon emission computed tomography (SPECT), and cardiac magnetic resonance (CMR). The aim was to extract ejection fraction and related measurements from report text.

Among 2,157 nuclear/CMR reports, 1,065 yielded an ejection fraction value. Manual review of 213 of those 1,065 reports, a 20% subset, found agreement for every reviewed extraction. That finding applies to the reviewed subset. For strain, review of 148 initially unresolved reports recovered 15 values; 133 contained no actual strain measurement.

The work demonstrates feasibility within one health system. Reviewers were unblinded, formal sensitivity, specificity and F1 estimates were not generated, and strain extraction lacked formal quantitative performance evaluation. Application elsewhere requires local adaptation and further validation. These limits matter when interpreting the results and planning subsequent research.

Explore the full methods and results

For Brown Virtual Lab: a milestone to learn from together

For our mentorship audience, this feature is an opportunity to make the research process approachable. We celebrate Dr. Brown and the publication team while inviting learners to notice the questions behind the finished article. A useful reading session might begin with a plain-language explanation of the research aim, followed by a discussion of how the authors connected clinical knowledge with computational rules.

Our suggested teaching approach is to ask learners to distinguish a reported result, an interpretation, and a proposed next step. Each serves a different purpose. Mentors can model how to speak confidently about a contribution while remaining precise about its limits. This is an editorial learning proposal, not a claim that the publication evaluated a mentorship program.

The celebration can also prompt a conversation about teamwork: how to ask for clarification across disciplines, document a decision, and acknowledge the people who helped resolve uncertainty. We hope readers carry that spirit of curiosity and shared credit into their own developing research practice.

Questions for discussion

1. Which sentence states the research question most clearly?
2. What would you ask a clinician before writing an extraction rule?
3. How would you explain the reviewed subset to a reader encountering validation for the first time?

Editorial perspective

making information usable

The following editorial reflections consider the meaning of the milestone and propose questions for future work. The published findings are presented in the study section above.

A clinical report can hold information that is valuable long after the encounter that produced it. Turning that information into a research dataset requires decisions about meaning, context, missingness, and consistency. We see this paper as an invitation to take that often quiet work seriously. The usefulness of an analysis begins with the care taken to understand its source material.

Continuing the conversation

recognition as an invitation

A publisher feature can open a door to a wider conversation. Researchers may approach the article through methods, learners through a first encounter with clinical informatics, and community readers through an interest in heart health after cancer. A thoughtful celebration gives each audience a place to begin. It pairs understandable language with direct access to the publication, so curiosity can lead to an informed reading.

We also welcome the opportunity to recognize collaboration. A publication carries the contributions of people with different kinds of expertise. Giving that work visible credit encourages a culture in which careful preparation, constructive review, and sustained cooperation are valued alongside the finished paper.

Continuing the conversation

keeping judgment visible

For future projects inspired by this topic, we would place the reasoning behind data decisions close to the data themselves. What does a missing value mean? Which statement in a report should take priority? When should an uncertain result return to a human reviewer? These are useful questions for project design, teaching, and evaluation.

Making such decisions visible can help collaborators understand what a dataset supports. It can also help a new team identify which assumptions must be revisited before using an approach in another setting. We regard that transparency as an important goal for future work, with its value to be examined through deliberate testing.

Continuing the conversation

a place for equity questions

Our editorial perspective also asks who is represented when historical records become research data. Future investigators could examine whether documentation, access to imaging, or missing measurements differ across the groups relevant to their question. The answers require appropriate data and analysis; they cannot be inferred from the existence of an automated extraction process.

These questions give equity a practical place in a research plan. They encourage teams to define whose experiences they hope to understand, assess the boundaries of available records, and explain those boundaries to readers. That is a constructive direction for continued inquiry.

Continuing the conversation

Read, discuss, and carry the questions forward

We invite readers to visit the Elsevier collection and the full article, examine the methods and limitations together, and bring a question to their next research discussion. The reflections below suggest starting points for different audiences. This milestone deserves celebration, and it offers an occasion to renew a shared commitment to clear questions, careful evidence, generous collaboration, and honest communication.

Publication record and team credit

This feature celebrates the original publication and its authors. The institute reflections are educational interpretations and proposed directions for future inquiry. Organizational sponsorship and author affiliations are documented in the original article.

Explore the evidence

• Elsevier World Heart Day article collection
• Original article on ScienceDirect
• PubMed bibliographic record
• Full article in PubMed Central
• World Heart Federation: World Heart Day
• Article reuse license: CC BY-NC-ND 4.0

Prepared September 30, 2026. Educational research communication. Clinical decisions require individualized professional assessment. The article retains its stated publication license; the visual on this page is original conceptual artwork.