Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Longitudinal analysis of CYFRA 21-1 levels in patients with pulmonary nodules: Differential trajectories between benign and malignant cases.

PloS one·2026
Same author

Impact of reconstruction kernel variability on segmentation consistency in low-dose thoracic CT.

Proceedings of SPIE--the International Society for Optical Engineering·2026
Same author

Personalized White Matter Bundle Segmentation for Early Childhood.

Proceedings of SPIE--the International Society for Optical Engineering·2026
Same author

Characterizing Continuous and Discrete Hybrid Latent Spaces for Structural Connectomes.

Proceedings of SPIE--the International Society for Optical Engineering·2026
Same author

Scalable Identification of Clinically Relevant Chronic Obstructive Pulmonary Disease Documents in Large-Scale Electronic Health Record Datasets With a Lightweight Natural Language Processing Model: Retrospective Cohort Study.

JMIR medical informatics·2026
Same author

Longitudinal Analysis of CYFRA 21-1 Levels in Patients with Pulmonary Nodules: Differential Trajectories Between Benign and Malignant Cases and Impact of Tumor Resection.

medRxiv : the preprint server for health sciences·2026
Same journal

What do LLMs value? An evaluation framework for revealing subjective trade-offs in assessment of glycemic control.

Proceedings of machine learning research·2026
Same journal

Towards the Efficient Inference by Incorporating Automated Computational Phenotypes under Covariate Shift.

Proceedings of machine learning research·2026
Same journal

Endo-SemiS: Towards Robust Semi-Supervised Image Segmentation for Endoscopic Video.

Proceedings of machine learning research·2026
Same journal

Perspective: Machine Learning for Health Should Consider Social Drivers of Health.

Proceedings of machine learning research·2026
Same journal

Classifying Phonotrauma Severity from Vocal Fold Images with Soft Ordinal Regression.

Proceedings of machine learning research·2026
Same journal

Does Domain-Specific Retrieval Augmented Generation Help LLMs Answer Consumer Health Questions?

Proceedings of machine learning research·2026
See all related articles

Related Experiment Video

Updated: Jun 14, 2026

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules
07:53

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

Contrastive Patient-level Pretraining Enables Longitudinal and Multimodal Fusion for Lung Cancer Risk Prediction.

Thomas Z Li1,2, Lianrui Zuo3, Yihao Liu3

  • 1Department of Biomedical Engineering, Vanderbilt University, Nashville, TN.

Proceedings of Machine Learning Research
|May 25, 2026
PubMed
Summary
This summary is machine-generated.

This study introduces a new method for predicting lung cancer risk using chest CT scans and clinical data. Our approach improves prediction accuracy by effectively combining information from different times and sources.

Keywords:
chest CTcontrastive language-image pretraining (CILP)lung cancermultimodal

More Related Videos

Unilateral Lung Volume Analysis Using Micro-CT for Enhanced Assessment of Pulmonary Fibrosis in Preclinical Models
03:38

Unilateral Lung Volume Analysis Using Micro-CT for Enhanced Assessment of Pulmonary Fibrosis in Preclinical Models

Published on: June 20, 2025

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections
06:22

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections

Published on: September 19, 2025

Related Experiment Videos

Last Updated: Jun 14, 2026

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules
07:53

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

Unilateral Lung Volume Analysis Using Micro-CT for Enhanced Assessment of Pulmonary Fibrosis in Preclinical Models
03:38

Unilateral Lung Volume Analysis Using Micro-CT for Enhanced Assessment of Pulmonary Fibrosis in Preclinical Models

Published on: June 20, 2025

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections
06:22

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections

Published on: September 19, 2025

Area of Science:

  • Artificial Intelligence
  • Medical Imaging
  • Clinical Informatics

Background:

  • Longitudinal and multimodal data are crucial for clinical prediction.
  • Contrastive Language-Image Pretraining (CLIP) excels with paired image-caption data.
  • Unpaired longitudinal data (e.g., CTs and notes at different times) presents a challenge.

Purpose of the Study:

  • To develop a patient-level contrastive pretraining method for longitudinal and multimodal data.
  • To address the challenge of aligning unpaired medical images and clinical variables.
  • To improve lung cancer risk prediction using this novel approach.

Main Methods:

  • Utilized a large public lung cancer screening dataset.
  • Employed a time-distanced transformer for longitudinal chest CT encoding.
  • Used open-source text embedding for clinical variable encoding.
  • Optimized contrastive loss between same-patient (positive) and different-patient (negative) pairs.

Main Results:

  • Finetuning CLIP representations significantly improved lung cancer risk prediction.
  • Achieved AUCs of 0.895 and 0.893 in two clinical populations.
  • Outperformed conventional multimodal fusion (0.873, 0.875 AUC) and single-modality baselines.
  • Demonstrated effective longitudinal and multimodal fusion without extra training data.

Conclusions:

  • Patient-level contrastive pretraining enables effective fusion of longitudinal and multimodal clinical data.
  • This method enhances lung cancer risk prediction accuracy.
  • The approach is valuable for leveraging ubiquitous, unpaired clinical data.