Introducing HOPPR™ AI Foundry 
 

The secure developer environment for building, fine-tuning, and validating medical imaging AI.   

The Foundry provides the secure infrastructure, trusted data, foundation models, and tools needed to accelerate AI development in medical imaging with speed and full traceability.   

Built for Builders. Trusted by Radiologists.  

Created by practicing radiologists and AI engineers, the HOPPR™ AI Foundry bridges technical scalability with the quality standards healthcare demands. 
 

 From data to deployment, every model, dataset, and workflow is transparent, auditable, and purpose-built to drive responsible innovation in healthcare. 

The HOPPR™ AI Foundry Includes

Intuitive User Interface (UI) 

An easy-to-use UI is layered over HOPPR’s APIs, lowering the barrier to entry for machine learning and medical imaging application developers.  

Fine-tune models without coding expertise, evaluate performance, and manage model versions without requiring custom compute environments or extensive DevOps support.  

The UI enables efficient experimentation, validation, and version tracking all within a secure, compliant platform.  

Foundation Models

HOPPR™ AI Foundry includes two large-scale Vision Transformer (ViT) foundation models, trained on tens of millions of labeled, annotated, and de-identified medical images, designed as a starting point for development. 

Based on self-supervised learning (SSL), these models learned robust visual representations from unlabeled data — generating pseudo-labels, uncovering hidden patterns, and generalizing across diverse datasets. 

They are balanced to support a diverse range of patient demographics, sites, and imaging systems. 

Results are delivered in structured formats, standardized JSON, and text for easy integration.  

HOPPR™ MC Chest Radiography
Foundation Model 

  • Trained on ~12.2M images / ~6.1M studies.

  • Supports binary classification fine-tuning with model score output 

  • Internal validation: Fine-tuning has yielded a median ROC-AUC of 0.91 across 24 findings (range 0.77–0.99), demonstrating potential for fine-tuning on downstream use cases.  

HOPPR™ EB 2D Mammography
Foundation Model 

  • Trained on ~24M images / ~4M studies 

  • Supports binary (e.g., cancer detection and pacemaker identification) 

  • Internal validation: Fine-tuning has shown ROC-AUC 0.9 (cancer), 0.94 (density), 0.99 (pacemaker) 

Dataset Highlights: 

  • Mammography: 5400 pathology-proven studies, includes laterality labeling (e.g., left vs. right breast) 

  • Chest Radiography: 25 datasets featuring 200, 500 unique studies for PA / AP projections

  • Granularity: Laterality and metadata provenance unique to HOPPR’s dataset design 

  • Consistency: Enables you to reproduce HOPPR’s baseline model performance

Flexible Data Options

You can bring your own imaging data or use HOPPR’s curated, labeled, and validated datasets to fine-tune foundation models for specific use cases. 

This flexibility supports experimentation, reproducibility, and faster iteration without compromising compliance.  

HOPPR Labeled and Validated Datasets

HOPPR™ AI Foundry provides one of the largest repositories of labeled and validated imaging data with verified provenance. These datasets are annotated, validated, and balanced with positive and negative examples for reliable training. 

Fine-tuning and Inference 

We make it easy for you to fine-tune and evaluate models to meet your specifications with our tools and compute. 


Then, you can run inference on your pre-trained model.  

Model Integration 

Deploy models into production-ready environments with control and flexibility.  

The HOPPR™ AI Foundry makes it simple to connect your fine-tuned or third-party models to your existing applications and workflows, ensuring your team can extend innovation securely and at scale.  

From our Foundry, you can easily integrate your fine-tuned or third-party models into your application using developer-friendly RESTful APIs.  

 Disclaimer: You are responsible for making any necessary modifications, validating model performance in the final product, and obtaining any applicable regulatory marketing authorizations before commercialization.

QMS-Aligned Framework within a Secure, Scalable Infrastructure 

The Foundry was developed under a quality management system (QMS) aligned with ISO 13485, IEC 62304, ISO/IEC 4200, and ISO 14971. This framework provides the quality, safety, and traceability foundation to support regulatory preparation and documentation generation throughout the development lifecycle. 

The Foundry operates within a secure, HIPAA-compliant environment supporting fine-tuning and inference via API, usage-based billing, and end-to-end traceability. 

HOPPR™ provides a compliant foundation designed specifically for ingesting, fine-tuning, validating, and hosting imaging models — enabling faster, responsible AI development at scale.

Accelerating Your Responsible AI Development 

The HOPPR™ AI Foundry was built to lower the barriers to responsible AI development in medical imaging. By uniting foundation models, curated datasets, and secure infrastructure, it empowers organizations to build imaging AI solutions faster, more efficiently, and with confidence in their compliance readiness.