Smart Pathology

Pathology Foundation Model Application Platform

A Vibe Research style, zero-code platform that lets hospitals build and validate specialty pathology AI models for long-tail tumors and research questions.

Smart PathologyPathology departments, clinical research teams and hospitals building proprietary pathology AI capability.
Pathology Foundation Model Application Platform

Core Capabilities

Designed around real clinical workflow.

  • Zero-code model development for clinician-led AI innovation.
  • End-to-end model production covering data, annotation, training, evaluation and release.
  • Small-sample modeling for specialty diseases and local research needs.
  • Standardized multi-metric model evaluation.
  • Hospital-led Vibe Research workflow: clinicians raise the question, AI supports model building, validation and iteration.
  • Pathology image lossless compression model capability for large-scale slide storage, research and model training.
  • PathEngine links slide management, semi-automatic annotation, training, evaluation and AI-assisted diagnosis feedback into one production loop.

Customer Value

What the product changes in healthcare workflows.

  • Turns hospitals from AI users into AI producers.
  • Accelerates dedicated models for specialty disease scenarios.
  • Shortens the path from data and research question to model validation.
  • Supports local training, validation and clinical application under hospital control.
  • Helps hospitals move from buying single AI modules to continuously producing models from their own clinical questions.

Deployment Proof

Evidence, outcomes and representative customers.

Explore how SenseCare supports clinical teams, from everyday workflows to deployment at scale.

Deployed in nearly 30 backbone hospitals.

Xijing Hospital used HE slides to predict key renal cancer driver mutations.

Independent test AUC reached 0.95 for BAP1 mutation prediction.

Shanghai Pulmonary Hospital completed frozen small-nodule subtype modeling in under 2 months.

12x-24x modeling efficiency improvement and 99.4% bronchiolar adenoma accuracy in product validation.

A global pathology lossless compression foundation model collaboration used 800+ representative WSI and 10M+ image patches for training and testing.

Shanghai Pulmonary Hospital launched cytology models for pleural and ascitic fluid, bronchial brushing, lung puncture and EBUS-TBNA workflows.

Shanghai Pulmonary Hospital's frozen small-nodule work completed 1,200+ high-quality annotations and 800+ training iterations for preliminary validation.

Representative Customers

Known deployments and collaboration references.

Xijing HospitalShanghai Pulmonary HospitalAnhui Provincial HospitalJiangsu Province HospitalXinhua Hospital