Bridging the Gap: From Machine Learning Research to Clinical Practice

Tue Dec 14 8:30am -- 5:30pm EST @ NeurIPS 2021 (Virtual)

Accepted Papers & Awards

Best Paper Awards:

  • Identification of Subgroups With Similar Benefits in Off-Policy Policy Evaluation
    Ramtin Keramati (Stanford University); Omer Gottesman (Brown University); Leo Celi (BIDMC); Finale Doshi-Velez (Harvard University); Emma Brunskill (Stanford University)

  • GAM Changer: Editing Generalized Additive Models with Interactive Visualization
    Zijie J Wang (Georgia Institute of Technology); Alex Kale (University of Washington); Harsha Nori (Microsoft Research); Peter Stella (NYU Langone Health); Mark Nunnally (NYU Langone Health); Duen Horng Chau (Georgia Institute of Technology); Mihaela Vorvoreanu (Microsoft Research); Jennifer Wortman Vaughan (Microsoft Research); Rich Caruana (Microsoft Research)

Best Reviewer Award:

  • Eugene Bykovets, ETH Zurich, Switzerland

  • Trenton Chang, University of Michigan, USA

Spotlights


  • GAM Changer: Editing Generalized Additive Models with Interactive Visualization
    Zijie J Wang (Georgia Institute of Technology); Alex Kale (University of Washington); Harsha Nori (Microsoft Research); Peter Stella (NYU Langone Health); Mark Nunnally (NYU Langone Health); Duen Horng Chau (Georgia Institute of Technology); Mihaela Vorvoreanu (Microsoft Research); Jennifer Wortman Vaughan (Microsoft Research); Rich Caruana (Microsoft Research)

  • Rethinking Generalization Performance of Surgical Phase Recognition with Expert-Generated Annotations
    Seungbum Hong (hutom); Jiwon Lee (hutom); Bokyung Park (hutom); Ahmed Abbas Alwusaibie (Yonsei University College of Medicine ); Anwar Hudaish Alfadhel (Yonsei University College of Medicine); SungHyun Park (hutom); Woo Jin Hyung (hutom); Min-Kook Choi (hutom)

  • Identification of Subgroups With Similar Benefits in Off-Policy Policy Evaluation
    Ramtin Keramati (Stanford University); Omer Gottesman (Brown University); Leo Celi (BIDMC); Finale Doshi-Velez (Harvard University); Emma Brunskill (Stanford University)


Posters


  • Survival-oriented Embeddings for Improving Accessibility to Complex Data Structures
    Tobias Weber (LMU Munich); Michael Ingrisch (LMU Munich); Matthias Fabritius (LMU Munich); Bernd Bischl (LMU Munich); David Ruegamer (LMU Munich)

  • What Do You See in this Patient? Behavioral Testing of Clinical NLP Models
    Betty van Aken (Beuth University of Applied Sciences Berlin); Sebastian Herrmann (Beuth University of Applied Sciences Berlin ); Alexander Löser (Beuth Hochschule für Technik Berlin)

  • Longitudinal Fairness with Censorship
    Wenbin Zhang (Carnegie Mellon University); Jeremy C Weiss (Carnegie Mellon University)

  • A Tree-based Model Averaging Approach for Personalized Treatment Effect Estimation from Heterogeneous Data Sources
    Xiaoqing Tan (University of Pittsburgh)

  • Interpretable Data Analysis for Bench-to-Bedside Research
    Zohreh Shams (University of Cambridge); Botty Dimanov (University of Cambridge); Sumaiyah Kola (University of Cambridge); Nikola Simidjievski (University of Cambridge ); Helena Andres Terre (University of Cambridge); Paul M Scherer (University of Cambridge); Urška Matjašec (University of Cambridge); Jean Abraham (University of Cambridge); Mateja Jamnik (University of Cambridge); Pietro Lió (University of Cambridge)

  • Transferring Multi-Omics Survival Models to Clinical Settings Through Linear Surrogate Models
    David N Wissel (ETH Zürich); Daniel Rowson (ETH Zürich); Valentina Boeva (ETH Zürich)

  • Contextualized Representation Learning in Biomedical Word Sense Disambiguation
    Mozhgan Saeidi (Dalhousie University)

  • Interpretable Electrocardiogram Mapping to Detect Decreased Cardiac Contraction
    Hirotoshi Takeuchi (University of Tokyo); Satoshi Kodera (University of Tokyo); Mitsuhiko Nakamoto (University of Tokyo); Shinnosuke Sawano (University of Tokyo); Susumu Katushika (University of Tokyo)

  • Designing A Clinically Applicable Deep Recurrent Model to Identify Neuropsychiatric Symptoms in People Living with Dementia Using In-Home Monitoring Data
    Francesca Palermo (Imperial College London); Honglin Li (Imperial College London); Alexander Capstick (Imperial College London); Nan Fletcher-Lloyd (Imperial College London); Yuhen Zhao (Imperial College London); Samaneh Kouchaki (Imperial College London); Ramin Nilforooshan (UK Dementia Research Institute); David Sharp (Imperial College London); Payam Barnaghi (Imperial College London)

  • Type Safety and Disambiguation of Depression
    Michael A Yee (EleutherAI)

  • Automated Supervised Feature Selection for Differentiated Patterns of Care
    Catherine Wanjiru (Carnegie Mellon University Africa); William Ogallo (IBM Research); Girmaw Abebe Tadesse (IBM Research); Charles Wachira (IBM Research); Isaiah Onando Mulang' (IBM Research - Africa); Aisha Walcott-Bryant (IBM Research - Africa)

  • Harmonizing Attention: Attention Map Consistency For Unsupervised Fine-Tuning
    Ali Mirzazadeh (Georgia Institute of Technology); Florian Dubost (Stanford University); Maxwell Pike (Stanford University); Krish Maniar (Stanford University); Daniel Y Fu (Stanford University); Khaled K Saab (Stanford University); Christopher Lee-Messer (Stanford University); Daniel Rubin

  • Post-discovery Analysis of Anomalous Subsets
    Isaiah Onando Mulang' (IBM Research - Africa); William Ogallo (IBM Research); Girmaw Abebe Tadesse (IBM Research); Aisha Walcott-Bryant (IBM Research - Africa)

  • Robust Interpretable Rule Learning to Identify Expertise Transfer Opportunities in Healthcare
    Willa Potosnak; Sebastian Caldas (Carnegie Mellon University); Keith Dufendach (UPMC); Giles Clermont (University of Pittsburgh); James K Miller (Carnegie Mellon University); Artur Dubrawski (Carnegie Mellon University)

  • Predicting Sufficiency for Hemorrhage Resuscitation Using Non-invasive Physiological Data without Reference to Personal Baselines
    Xinyu Li (Carnegie Mellon University); Michael R Pinsky (University of Pittsburgh); Artur Dubrawski (Carnegie Mellon University)

  • Neuroweaver: Towards a Platform for Designing Translatable Intelligent Closed-loop Neuromodulation Systems
    Parisa Sarikhani (Emory University); Hao-Lun Hsu (Georgia Institute of Technology); Joon Kyung Kim (University of California, San Diego); Sean Kinzer (University of California, San Diego); Edwin Mascarenhas (University of California, San Diego); Hadi Esmaeilzadeh (University of California, San Diego); Babak Mahmoudi (Emory University)

  • A Conservative Q-Learning approach for handling distributional shift in sepsis treatment strategies
    Pramod Kaushik (International Institute of Information Technology Hyderabad); Sneha Kummetha (Independent Researcher); Perusha Moodley (Independent Researcher); Raju Surampudi Bapi (International Institute of Information Technology Hyderabad)



GlaxoSmithKline (GSK) is a science-led global healthcare company with a special purpose to improve the quality of human life by helping people do more, feel better, live longer. Every day, we help improve the health of millions of people around the world by discovering, developing and manufacturing innovative medicines, vaccines and consumer healthcare products. We are building a stronger purpose and performance culture underpinned by our values and expectations - so that together we can deliver extraordinary impact for patients and consumers.

GSK uses AI to discover transformational medicines. AI is the key to interpret genetic datasets so we can understand the 'language' of the cell and develop medicines with a higher probability of success.


We are pleased to support this award for the best paper, describing innovative machine learning research focused on bridging the gap between machine learning research and application in clinical practice.

Note:

  • GSK is not involved in the selecting the winning paper.

  • To claim the award, the lead author of the paper selected will be required to complete the appropriate brief documentation in accordance with healthcare regulatory best practice and compliance requirements.