AI Startup Joins Radiology Giant to Scale Healthcare Access

Louis Blankemeier and his co-founders built something remarkable during their Stanford PhDs: AI that reads X-rays and CT scans like radiologists do, generating comprehensive medical reports across thousands of possible diagnoses. In October 2024, they launched Cognita to bring this technology to hospitals and clinics.

Less than a year later, they faced every founder's dream dilemma: accept venture capital to stay independent, or sell to Radiology Partners, the world's largest radiology practice. They chose the path Silicon Valley usually mocks: they sold.

Their reasoning turns conventional startup wisdom upside down. In healthcare, brilliant technology isn't enough. Clinical AI requires massive datasets, continuous real-world testing, regulatory clearances, and trust built through proven safety records.

The team realized something crucial: their research models, trained on tens of thousands of scans, weren't ready for real patient care. A single CT scan can contain a billion pixels encoding entire medical textbooks of information. Edge cases involving rare but critical conditions appear regularly in clinical practice.

Blankemeier points to self-driving cars as a parallel. After a decade and billions invested, only companies that controlled the entire system achieved reliability: the vehicles, sensors, data collection, and deployment infrastructure all working together at massive scale.

Radiology demands the same integration. Success requires continuous data feeds surfacing rare cases, vast clinical resources redesigning workflows, regulatory expertise, and constant safety monitoring. Most importantly, it needs something nearly impossible for startups to access: feedback from radiologists reviewing and correcting AI-drafted reports.

The Ripple Effect

That correction data creates a powerful cycle. Better AI models help radiologists work more accurately and handle more patients. Higher accuracy improves future training data. Increased capacity lets radiology groups take more contracts, generating even more data and corrections.

For a startup, building this flywheel independently would take years of convincing skeptical hospitals to trust unproven technology. By joining Radiology Partners, Cognita immediately gained access to diverse patient populations, multiple medical centers, and the credibility needed to expand quickly.

In healthcare, evidence isn't built through small pilots. Trust comes from sustained performance across different sites, patient types, and medical scenarios. Operating within the world's largest radiology practice lets them prove efficacy, reliability, and security simultaneously.

The founders didn't give up their ambition. They protected their velocity by structuring the deal to maintain their mission: significantly increasing global healthcare access through AI technology that actually works in real clinical settings.

Sometimes the fastest way forward means joining forces with those already doing the work at scale.