AI Pathology Targets Skin Cancer with Nvidia Support

A Brazilian healthtech firm, AI Pathology, is advancing its mission to combat skin cancer through artificial intelligence, bolstered by a strategic partnership with Nvidia. The company is leveraging Nvidia's Inception program, which provides access to crucial computing resources, expert guidance, and training, to accelerate its development of AI-driven diagnostic tools.

Founded in 2024 by dentists Willian Boelcke and Lucas Souza, AI Pathology aims to create the world's most extensive image repository of skin lesions. This ambitious goal is designed to harness Brazil's significant skin tone diversity as a competitive advantage, enabling the development of AI models with global applicability. Currently, the startup has amassed approximately 90,000 images through collaborations with institutions across Brazil, with a target of reaching 4 million images to build a robust dataset for dermatological AI training.

The company's flagship application, 'Nevo', allows users to capture images of skin lesions via smartphone, receiving an immediate risk assessment. This accessible technology, functional on basic mobile devices, boasts an impressive 93% accuracy rate. Skin cancer remains the most prevalent cancer in Brazil, with an estimated 230,000 new cases annually, underscoring the critical need for accessible early detection solutions.

A significant pilot program conducted in partnership with Senar in Goiás demonstrated the technology's impact. The initiative screened 2,058 rural workers, a demographic highly susceptible to sun exposure. The AI identified approximately 10% of individuals requiring priority assessment and detected 75 early-stage skin cancer cases, facilitating prompt medical intervention. This project garnered recognition at a Harvard School of Public Health Hackathon and led to AI Pathology's inclusion in the Harvard Innovation Labs.

Willian Boelcke believes AI Pathology's model can alleviate the strain on Brazil's healthcare system, particularly the lengthy waiting lists for dermatologists. With a national ratio of roughly one dermatologist per 17,000 inhabitants, and a concentration in wealthier regions, wait times can extend up to three years in underserved areas. Nvidia's Márcio Aguiar highlighted that AI in medicine serves to augment, not replace, physician decision-making, a principle AI Pathology embodies.

Having secured R$ 4 million from Enseada, the family office of the SulAmérica founders, AI Pathology is preparing for regulatory approval from Anvisa, which began regulating AI medical software in 2024. Upon receiving clearance, the company plans to launch a Series A funding round targeting approximately $8 million to fuel international expansion. This includes clinical studies in the United States, with the aim of securing FDA approval, a process where AI Pathology anticipates its diverse dataset will be a key differentiator, potentially making it one of the first companies to meet the FDA's demographic representativeness requirements for AI health tools.

The company's business model is primarily B2B, with existing partnerships including Natura, which offers the technology as an employee benefit, and the laboratory Sabin, integrating it into its diagnostic network. AI Pathology's journey, originating from a personal tragedy, is driven by a commitment to preventing similar outcomes through accessible and advanced AI diagnostics.