ViewsML Builds Computational Layer for Virtual Biomarker Diagnostics

Kenneth To and his team at ViewsML are tackling one of the persistent, high-friction bottlenecks in modern precision medicine: the sheer labor and cost associated with traditional tissue analysis. Their ambition is to transform the analysis of human tissue samples—the bedrock of diagnostic pathology—from a specialized, physical lab process into scalable, software-driven computation. This is not simply digitizing an existing workflow; it's building what they term 'the computational layer for next-generation diagnostics.'

At its core, ViewsML has engineered a platform capable of virtual immunohistochemistry (vIHC). Historically, identifying a biomarker—a specific molecular marker linked to a disease state—requires laborious chemical staining (like IHC) applied to a physical tissue slide. This process is slow, expensive, and can often destroy the very sample needed for follow-up testing. ViewsML’s solution uses advanced AI models to extract detailed, per-cell biomarker spatial and quantitative insights directly from a routine, standard Hematoxylin and Eosin (H&E) slide. This capability allows researchers to 'see' biomarker staining virtually, bypassing the traditional need for time-consuming chemical stain batches and significantly accelerating the time-to-insight from weeks to minutes.