MLHC 2023 : Machine Learning for Healthcare
Call For Papers
MLHC has two tracks a research track, for full papers to be archived, and a clinical abstract track, non-archival short papers highlighting work by clinicians as well as software and demos. Please read the instructions about each track (below) carefully to determine which track best fits your work.
IMPORTANT 2023 DATES:
Paper Submission Deadline — April 12, 2023
Author Response Due — May 25, 2023
Acceptance Notification — June 21, 2023
Research Track: Call for Papers, Review Process
CALL FOR PAPERS:
For decades, researchers in computer science and medical informatics have developed and applied machine learning techniques in the hope of leveraging data to derive insights that could advance clinical medicine. Recently, advances in machine learning (spanning theory, methods, and tooling) and digital medicine (the advent of EHRs, public datasets, and technologically minded clinicians) have created ideal conditions for leaps forward in machine learning for healthcare. To realize this challenge, however, we must tackle two grand challenges:
(i) leveraging complex data (images, other sensor data, and patient records consisting both raw and unstructured data captured at irregular intervals);
(ii) the need to provide actionable insights (such as robust causal inferences about the likely impacts of interventions).
Moreover, realizing the potential of machine learning in healthcare requires that technical researchers and clinicians work together to identify the right problems, obtain the right data, and verify the conclusions, and ultimately realize the potential of proposed solutions in practice. While advances in deep learning have made a dent on the complex data front, there’s far more work to be done. Meanwhile, the leap from prediction to decision-making remains in its infancy. The Machine Learning for Healthcare Conference (MLHC) is the premier publishing venue solely dedicated to work at this vibrant intersection. MLHC has brought thousands of machine learning and clinicians researchers together since its inception to present groundbreaking work (archived in the Proceedings of Machine Learning Research) and to forge new collaborations. We hope that you will submit your strongest work to MLHC 2023 and will join us at Columbia in August for the conference.
Appropriate submissions include both (i) novel methods that tackle fundamental problems arising in healthcare data (including sparsity, multimodal data, class imbalance, temporal dynamics, distribution shift across populations, and the need to estimate treatment effects); and (ii) end-to-end machine learning solutions to important problems in healthcare (including new methods, insightful evaluations of existing methods with results of interest to the community, and in-vivo analyses of systems deployed in the wild). We also welcome replication studies - please contact the organizers prior to submission to ensure that your paper is within scope and reviewed under the appropriate track. However, survey papers which simply summarize existing methods will not be accepted. Submissions will be reviewed by both computer scientists and clinicians. This year, like previous years, we are calling for papers in two tracks: a research paper track and a clinical abstract+software/demo track. Accepted papers will be archived through the Proceedings of Machine Learning Research (JMLR Proceedings track).
While it’s impossible to enumerate every conceivable problem of interest, our guiding principle is that accepted papers should provide important new generalizable insights about machine learning in the context of healthcare.
Additional Context for Clinicians: We realize that conferences in medicine tend to be abstract-only, non-archival events. This is not the case for MLHC: to be a premier health and machine learning venue, all papers submitted to MLHC will be rigorously peer-reviewed for scientific quality -- and for that a suitably complete description of the work is necessary. So we call for submissions that describe your problem, cohort, features used, methods, results, etc. Multiple reviewers will provide feedback on the submission. If accepted, you will have the opportunity to revise the paper before submitting the final version. If you wish to submit a shorter, non-archival paper, see the Clinical Abstracts track below.
Additional Context for Computer Scientists: MLHC is a machine learning conference, and we expect papers of the same level of quality as those that would be sent to a conference (rather than a workshop).
Please use the full paper LaTeX files available [here]. The example paper in the file pack contains sample sections. A more machine-learning oriented paper may include more mathematical details, while a more application-focused paper may include more detailed cohort and study design descriptions. In all cases, papers should contain enough information for the readers to understand and reproduce the results. Moreover, you must keep the generalizable insights heading in the introduction. The margins and author block must remain the same and all papers must be in 11-point Times font.
We expect papers to be between 10-15 pages (excluding references). While there is no strict page limit, the appropriateness of additional pages beyond the recommended length will be judged by reviewers. Please refer to the submission instructions on our website, including tips on what makes a great MLHC paper and required content.
Finally, papers must be submitted blinded. Do not include your names, your institution’s name, or identifying information in the initial submission. Wait for the camera-ready. While you should make every effort to anonymize your work -- e.g., write “In Doe et al. (2011), the authors…” rather than “In our previous work (Doe et al., 2011), we…” -- we realize that a reviewer may be able to deduce the authors’ identities based on the previous publications or technical reports on the web. This will not be considered a violation of the double-blind reviewing policy on the author’s part.
All papers will be rigorously peer-reviewed by both clinicians and ML researchers, with an emphasis on what generalizable insights the work provides about machine learning in the context of healthcare. We encourage you to read the content on “How to Write a Great MLHC Paper.” to see what we are looking for as well as common pitfalls. Reviewing for MLHC is double-blind: the reviewers will not know the authors’ identity and the authors will not know the reviewers’ identity (see format section above for more notes).
Finally, below are the questions on the review form, so that you know exactly what reviewers will be asked about your paper:
Summarize the paper and its main contributions
What generalizable insights did the authors claim they are making to machine learning in the context of healthcare?
Were the claims of these insights supported in the body of the paper?
Please provide detailed comments, including strengths and weaknesses of the paper.
Is your main expertise on the clinical or computational side (or both)?
PROCEEDINGS AND PRESENTATIONS:
Accepted submissions will be published through the Proceedings of Machine Learning Research (formerly the JMLR Workshop and Proceedings Track) and we are in the process of securing indexing on PubMed. Publications through JMLR are made open access without an article processing fee. Authors of accepted papers will be invited to present a spotlight and/or a poster on their work at the conference.
It is a violation of dual-submission policy to publish at MLHC and then later submit the same paper to another conference.
DUAL SUBMISSION POLICY:
Concerning dual submissions, research that has previously been published, or is under review, for an archival publication elsewhere may not be submitted. This prohibition concerns only archival publications/submissions and does not preclude papers accepted or submitted to non-archival workshops or preprints (e.g., to arXiv).
Clinical Abstract Track
CALL FOR ABSTRACTS:
In addition to our main research proceedings, we welcome the submission of both (i) clinical abstracts; and (ii) software/demo abstracts, to be presented in a non-archival track: Clinical abstracts typically pitch clinical problems ripe for machine learning advances or describe translational achievements. The first author and presenter of a clinical abstract track submission must be a clinician (often an MD or RN). Software demos typically introduce a tool of interest to machine learning researchers and/or clinicians in the community to use. These are often (but not necessarily) open source tools. The abstract may consist of:
Open clinical questions or interesting data sets: we seek viewpoints from clinicians and clinical researchers on important directions the MLHC community should tackle together, as well as abstracts describing interesting data sources.
Preliminary computational results: we encourage submissions from clinical researchers working with digital health data using modern computational methods; MLHC is a great venue for clinical researchers to brainstorm further analyses with an engaged computational community.
Clinical/translational successes: we seek abstracts about data and data analysis that resulted in new understanding and/or changes in clinical practice.
Demonstrations: we seek exciting end-to-end tools that bring data and data analysis to the clinician/bedside.
Software: we seek abstracts describing processing tools/pipelines tailored to health data.
Abstracts will not be archived or indexed, but will have the opportunity to be presented as a poster and/or spotlight talk at MLHC. Given that this track is designed to engage clinicians, the first author of a clinical abstract must be a clinician (MD, RN, etc. -) does your job involve working with patients) or a clinician-in-training (i.e., currently enrolled in an MD or MD/PhD program). The clinical abstract track is not intended for work-in-progress by primarily computational researchers.
Submissions should be one page or less, using the abstract template [linked here]. Clinical abstracts are not blinded; author names, degrees, and affiliations should be present in the submission.
All clinical abstracts will be peer-reviewed.
PROCEEDINGS AND PRESENTATIONS:
Abstracts will not be archived. Authors of accepted papers will be invited to present a spotlight and/or a poster on their work at the conference. One of the presenting authors must be a clinician.
DUAL SUBMISSION POLICY:
Work in progress, work in submission, and recently published work are all welcome (as long as you follow the other publication’s rules).