HOPPR

HOPPR offers a multimodal model named “Grace” for medical imaging, catering to researchers and developers through a private beta phase for early access (as of December 2023), enabling fine-tuning and application development for medical purposes. It introduces interactive imaging, enabling healthcare professionals to engage with imaging studies, aiding in diagnosis and decision-making. For instance, radiologists can benefit from automated report generation and interactive sessions with imaging studies via API, while neurosurgeons and trauma surgeons can leverage predictive tools for planning procedures and prioritizing high-risk patients, respectively. Moreover, the platform targets academic medical centers and medical office staff, offering automation in research tasks and improving billing accuracy to maximize revenue for medical facilities.

The platform, built on AWS, offers scalability and security, complying with privacy regulations such as HIPAA. It allows users’ control over data access, offering a clean development environment with pre-loaded tool sets like TensorFlow for quick setup. Additionally, it supports ingestion, anonymization, and indexing of large datasets for multi-center AI model development. HOPPR includes a proprietary search tool for building specific cohorts for analysis, leveraging open-source and proprietary tools specialized for medical imaging. The platform possesses diverse medical imaging datasets covering MRI, CT, Echo, and X-ray with studies, images, and reports.

Key customers and partnerships

In September 2024, HOPPR partnered with RadNet's DeepHealth, an AI-powered health informatics company, to advance AI applications in healthcare.