Kayode Odunuga, Medical Doctor at the Olabisi Onabanjo University Teaching Hospital, shared a post on LinkedIn:
“Honored to deliver an Oral Presentation at the National Institute for Cancer Research and Treatment!
I took the stage to showcase ICW023 – OncoTrackka: A Groq-Powered AI Platform for Enhanced Oncology Diagnosis and Patient Connectivity.
Sharing our vision for faster, smarter, and more connected cancer care with the nation’s leading experts was an absolute privilege!
The Challenge
Cancer diagnosis and patient-oncologist communication remain major hurdles in Nigeria’s healthcare ecosystem. Delays in interpretation of results and fragmented follow-up care affect outcomes.
Our Solution: OncoTrackka
We built OncoTrackka, a full-stack web platform that combines:
- Groq’s blazing-fast AI inference
- Meta’s LLaMA 3 for natural language understanding of multimodal oncology data
- Supabase for secure, HIPAA-compliant data handling
- Built-in telemedicine & patient-oncologist connectivity features
Key Findings
Groq’s acceleration delivered real-time AI responses – critical for time-sensitive oncology decisions
Diagnostic accuracy was low since LLaMA 3 wasn’t fine-tuned on Nigeria-specific medical data
BUT – the platform correctly flagged high-risk cases and recommended further specialist review
Conclusion and The Road Ahead
OncoTrackka can meaningfully bridge gaps in oncology care—even when starting with general-purpose models. That said, a persistent challenge in Nigeria’s health-tech landscape remains access to quality datasets. Cancer-specific data is either unavailable, heavily gated behind bureaucracy, or fragmented across institutions.
Fine-tuning on locally relevant, multimodal Nigerian oncology datasets would dramatically boost diagnostic accuracy and enable us to pivot toward AI-powered image analysis for histopathology and radiology—unlocking truly precision-grade insights.
We’re more convinced than ever that the future of scalable cancer care in Nigeria lies at the intersection of cutting-edge compute, open collaboration, and liberated data.
We’re actively seeking partners who can help us overcome these barriers—whether through data-sharing agreements, institutional collaborations, or policy advocacy.
If you have access to Nigerian oncology datasets, imaging archives, or simply the influence to open doors, let’s talk. Together, we can turn today’s proof-of-concept into tomorrow’s life-saving standard.”
More posts featuring Kayode Odunuga.