Nucs AI and Segmed are linking AI with a real-world oncology dataset spanning 2,800+ care sites and 150 million imaging studies to boost cancer prediction in precision oncology. The partnership aims to improve response prediction by pairing imaging with clinical outcomes, building on Nucs AI models already trained on more than 72,000 tumour lesions and addressing the generalizability problem seen with smaller datasets.

Why It Matters To Oncology

• Precision oncology models often underperform when trained on narrow, single-site data that capture local practice patterns rather than disease biology.

• By combining multi-institutional imaging with outcomes data, the collaboration could support more accurate predictive biomarkers, patient stratification tools and treatment-response models across cancer types.

• The companies cited the 2021 VISION trial as a reminder of the need: more than half of patients treated with Novartis' Pluvicto failed to achieve a PSA response.

The Financials

• Segmed will provide Nucs AI access to its oncology dataset, which covers multiple geographies, modalities and treatment settings.

• The deal also includes a strategic investment by Segmed in Nucs AI, though financial terms were not disclosed.

• Nucs AI recently also partnered with AstraZeneca to develop AI-driven response prediction models for therapeutic radioconjugates in metastatic prostate cancer.

What They're Saying

• "Models are not the bottleneck in precision medicine; data is," said Nucs AI CEO Nijat Ahmadov.

• Ahmadov said pairing imaging with real clinical context and outcomes can "unlock accuracy and patient-level specificity that imaging alone simply leaves on the table."

What's Next

• The partners plan to develop predictive biomarkers, clinical decision tools, disease-specific registries, research cohorts, real-world evidence programs and validation frameworks.

• Nucs AI is expected to expand into additional cancer types, imaging modalities and clinical applications using Segmed's dataset.

• If validated, the approach could strengthen AI use in oncology drug discovery, trial enrichment and treatment selection.