Skip to content

Understandable AI Imaging Biomarkers for Precision Oncology

Picture Health is developing an extensive pipeline of biologically interpretable, AI-driven image analysis algorithms. Each algorithm focuses on a deeper characterization of important anatomical and cellular features that directly inform a patient’s prognosis and likely response to different therapies.

Custom Biomarkers

Our Px™ Platform enables the development of custom biomarkers,  aligned with therapeutic mechanisms and trial goals. 

With the Px™ Platform, you can:

  • Build custom signatures that integrate radiology and/or pathology with clinical data
  • Utilize predictive and prognostic algorithms from pretreatment scans
  • Correlate with molecular biomarkers, including RNA expression (RNAseq) or genomic endpoints (Somatic NGS) for additional mechanisms of action (MoA) insights
  • Analyze pre/post scan analyses to quantify response in terms that align with treatment MoA

This allows sponsors to track treatment effects in ways traditional biomarkers cannot. 

 

custom

Standard Biomarkers

Our standard biomarkers are fully automated, non-invasive, and interpretable.

QVT Score

The first and only biomarker to quantify tumor vascularity and angiogenesis from routine imaging scans. 

  • Monitor on-treatment changes in tumor vascularity 
  • Predict the utility for use of anti-angiogenic therapies in the neoadjuvant and adjuvant setting
  • Validated across multiple checkpoint inhibitor and VEGF regimens
  • Built on routine imaging and independent of RECIST and tumor volume

CheckpointPx

A multimodal AI biomarker that expands beyond PD-L1 to predict outcomes to ICI regimens for patients with advanced NSCLC.

  • Integrate any available combination of routine radiology, pathology, or clinical data
  • Outperforms PD-L1 and RECIST measurements
  • Enables early OS signal for patients receiving standard-of-care ICI regimens