CVAI Catalog

·

View Catalog

tune

2 models found

·

2 public code

·

2 public weights

EchoCLIP

Cedars-Sinai Medical Center (Smidt Heart Institute) / Ouyang Lab · 2024

graph_1

Code & model weights public

Vision-language foundation model fine-tuned from CLIP on more than one million private echocardiogram video-report pairs, enabling zero-shot cardiac function assessment, device identification, and image/text retrieval without task-specific training. Combines a ConvNeXt-Base video encoder with a GPT-2-style text encoder under contrastive pretraining. Training data is private, but model weights and code are public. Developed by Cedars-Sinai's Ouyang lab.

Echocardiography video

Filter by Modality:
Echocardiography

Clinical text

Filter by Modality:
Text & EHR

Multimodal

Filter by Modality:
Multimodal

LVEF estimation

Filter by Disease / Trait:
Cardiac Function & Hemodynamics

General Purpose / Multi-task

Filter by Disease / Trait:
General / Foundation

Regression

Filter by Task Type:
Regression

Embedding

Filter by Task Type:
Representation Learning

Hybrid

Filter by Architecture:
Hybrid / Multi-branch

PyTorch

Filter by Framework:
PyTorch

Research use only

Filter by License:
Non-commercial / Research-only


Model ID: 0035

PanEcho

Yale School of Medicine (CarDS Lab) · 2025

graph_1

Code & model weights public

View-agnostic, multi-task model that performs 39 different echocardiographic reporting tasks - covering chamber size and function, valve disease, and more - from any combination of views, aggregating clip-level predictions up to the study level. Combines a ConvNeXt-Tiny frame encoder with a temporal Transformer and separate output heads per task. Trained on private Yale-New Haven Health System echo videos and published in JAMA in 2025 by Yale's CarDS Lab.

Echocardiography video

Filter by Modality:
Echocardiography

LVEF estimation

Filter by Disease / Trait:
Cardiac Function & Hemodynamics

LV dilation

Filter by Disease / Trait:
Structural Heart & Cardiomyopathy

LV systolic dysfunction (LVSD)

Filter by Disease / Trait:
Cardiac Function & Hemodynamics

Valvular disease

Filter by Disease / Trait:
Structural Heart & Cardiomyopathy

Structural heart disease (composite)

Filter by Disease / Trait:
Structural Heart & Cardiomyopathy

Regression

Filter by Task Type:
Regression

Binary classification

Filter by Task Type:
Classification

Multi-label classification

Filter by Task Type:
Classification

Hybrid

Filter by Architecture:
Hybrid / Multi-branch

PyTorch

Filter by Framework:
PyTorch

CC BY-NC-SA 4.0

Filter by License:
Non-commercial / Research-only


Model ID: 0043

·

Subject Count: 24,405