Bringing an AI Oral Cancer Test to Patients

Early detection of oral cancer saves lives. Machine learning applied to cytological images can flag lesions likely to become cancerous, but turning a university research tool into a clinical diagnostic requires validation, regulatory work, quality systems and a commercialization plan. 

A company developing a university-developed early oral cancer detection product wanted to move it from research tool to clinical diagnostic. 

The company needed to:

• Turn an ML-driven cytological image analysis tool into a clinical diagnostic. 
• Meet regulatory and quality requirements for a medical device. 
• Build the software, hardware and commercialization foundations for clinical use. 

Sia provided end-to-end product, regulatory and quality support: 

• Clinical lab: clinical lab software architecture, validation, a commercialization plan and a laboratory information management system for clinical testing. 
• Human factors: human factors studies and clinician report formats for fast, accurate results. 
• Regulatory: a breakthrough device submission and pre-submissions to the U.S. regulator, with medical device SOPs and traceability. 
• Product engineering: hardware, embedded software, edge and cloud AI/ML software, and user interface design and testing. 

Lives were already being saved: 

• The test launched as a clinical lab service in the United States, with the United Kingdom to follow, and the regulatory submission advanced. 
• Global commercialization was on track, with in-office results in 15 minutes. 

By combining product engineering, regulatory expertise and quality systems, Sia helps life sciences innovators bring new diagnostics to patients.