Home/Services/Clinical

Research phase 02

Clinical

Clinical samples arrive in batches, over months, from multiple sites. The hardest part is not measuring them — it is measuring the last batch exactly the way you measured the first. That is what this service is built around.

Consistency across every batch

Our image analysis scientists work across whole-slide images in a detailed and unbiased manner, so the data represents the staining and features actually present rather than the areas someone chose to look at. Algorithms are verified against ground truth and can then be locked, guaranteeing identical quantification across multiple sample batches.

  • Clinical pathologist input to support algorithm development, region-of-interest selection and lead annotation on clinical tissue images
  • Blinded analysis where the study design calls for it
  • Version-controlled algorithms with a documented change history
  • Batch-level QC reporting so drift is caught at the point it happens

Transparent from slide to table

  1. QC 1Image quality control

    Focus, staining intensity, tissue integrity and artefact checks on every image. Failures are reported with images attached, not silently dropped.

  2. QC 2Tissue annotation

    Regions of interest are annotated and reviewed. Where the anatomy is contentious, your pathologist arbitrates before analysis, not after.

  3. QC 3Algorithm development and lock

    Performance is verified on a held-out set against pathologist ground truth, signed off, then locked for the duration of the study.

  4. RUNBatch analysis

    Batches are processed as they arrive, each with its own QC record traceable back to the locked algorithm version.

  5. OUTReporting

    Interim and final data packages, with image sharing available for both Indica HALO and Visiopharm environments so your team can inspect the analysis directly.

Predicting response, defining populations

Quantification of specific tissue markers, mechanistic readouts or intercellular signalling proteins can indicate drug response or disease status within a defined patient population. Detailed measurement of how marker expression changes — across whole tissues or within specific regions — feeds directly into decisions about dose selection and trial inclusion criteria.

Use case

Biomarker translation

Carrying a pre-clinical readout into human tissue and checking it still separates responders from non-responders.

Use case

Stratification

Defining an expression cut-off that identifies the population most likely to benefit, with the distribution data to justify it.

Use case

Pharmacodynamics

Showing target engagement in tissue at a given dose, on paired pre- and post-treatment biopsies.

Bringing a biomarker into the clinic?

We will review the assay, the scanner output and the endpoint definition together, and tell you where the consistency risk sits before the first batch arrives.