HOPPR Unveils Chest Radiography Model with API Access

HOPPR has released its Marie Curie Chest Radiography Model, providing developers with fine-tuning and inference API access for binary classification of chest X-ray images, announced in a press release. This marks HOPPR's first commercial launch, aimed at enhancing developer infrastructure for medical imaging AI models.

The model, built on a Vision Transformer (ViT) architecture, allows developers to fine-tune using their own labeled datasets. It offers structured outputs and prediction scores via a single API. DeepHealth, a subsidiary of RadNet, is the first commercial partner to adopt this model, demonstrating its adaptability and potential in real-world applications.

HOPPR's platform provides a secure environment for customizing foundation models, supporting traceable and reproducible development workflows. The release includes an integrated billing portal, enabling teams to move from labeled data to deployment efficiently. Future updates will expand access to proprietary data and model validation capabilities, broadening support for customers lacking existing data or infrastructure.