Session
CSI-—
Model
CanScan v4.2.1
Ptr
0.000 · 0.000
Status
● OPERATIONAL
PRE-SEED

Reading
the invisible
signal
inside cancer.

[ Cytoscan Intelligence ]

Imaging-native oncology AI. We turn every CT, MRI and PET into a quantitative readout of the disease — thousands of features the human eye can't see, fused with molecular data to stage, stratify and predict response.

Scroll
AI-native cytology Fewer missed calls Clinical-grade accuracy Population scale
01  / MISSION
Precision is a measurement problem.
N=01 · REV. 04.18.2026

Every scan already contains the answer. Twelve-thousand quantitative features — texture, morphology, heterogeneity, perfusion — that a radiologist, trained on human vision, can only approximate. We built CanScan to read all of them, at once, and fuse them with the molecular state of the tumor.

Read the full story
Radiomic features
12,384/ scan
Per-voxel, multi-scale, across 4 modalities.
AUC lift vs. standard
+0.17mean
Across 4 solid-tumor cohorts, n = 8,412.
Earlier detection
14mo. lead
Lung adenocarcinoma, retrospective, NLST-follow.
Deployment
<90sper study
On-prem GPU or HIPAA-cleared cloud endpoint.
02  / PLATFORM
The CanScan architecture.
HIERARCHICAL ATTENTION FUSION NETWORK

A model that sees like a tumor grows.

Cancer is hierarchical — voxel, lesion, organ, patient — and so is CanScan. Six attention stages learn structure at every scale, then fuse imaging with proteomic signal through a shared latent space. No pixel is treated in isolation; no biomarker arrives without context.

Trained on 2.1 million studies across 34 institutions, validated prospectively in four solid-tumor cohorts. Deployable as a DICOM-in / JSON-out service in under 90 seconds per study.

DICOM · HL7 · FHIR HIPAA · SOC 2 · ISO 13485 ONNX · TensorRT
CanScan · architecture map ● LIVE
L01
Voxel Encoder3D patch tokenizer · 2M params
L02
Local AttentionSpatial · intra-tumor heterogeneity
L03
Regional AttentionInter-lesion · metastatic graph
L04
Cross-modal AttentionCT ↔ MRI ↔ PET fusion
L05
Molecular AttentionImaging ↔ proteomic embedding
L06
Hierarchical ReadoutStage · stratify · predict
DICOMin
Tokenizevoxels
Attend6 levels
Fusemodalities
ReadoutJSON
03  / IN OPERATION
From DICOM to decision in seconds.
SCROLL TO ADVANCE · CASE-2461·LUNG
PHASE · [ INGEST ]
04  / INDICATIONS
Four tumor types. More in trial.
EXPANSION ROADMAP · 2026–2027
● IN TRIAL
NSCLC · SCLC

Lung

Screening, staging, response to ICI. Trained on 580k thoracic studies across 14 sites.

AUC
0.94
Lead time
14mo
● IN TRIAL
PDAC

Pancreatic

The hardest cancer to catch early — we detect sub-centimeter lesions in dual-phase CT.

AUC
0.89
Sens. @ 1cm
82%
● IN DEVELOPMENT
ccRCC · papRCC

Renal

Non-invasive subtyping and risk stratification from contrast-enhanced CT/MRI.

AUC
0.91
Subtypes
5
● IN DEVELOPMENT
EC

Endometrial

Pre-operative risk stratification from pelvic MRI; myometrial invasion scoring.

Trial
n=812
Launch
Q3'26
05  / FUSION
Imaging × molecular, in a shared latent.
PROTEOMICS · GENOMICS · TRANSCRIPTOMICS

What a scan knows about a tumor's biology.

Radiomic features encode structure; proteomic assays encode function. CanScan learns a joint embedding across both, so a CT can predict protein expression, and a blood draw can refine a scan. The result is a single, continuous readout of the tumor — not a pile of disconnected reports.

This powers three clinical jobs: earlier detection, cleaner stratification, and response prediction — the three places oncology outcomes are won or lost.

Latent space · 1,024-dim

Imaging
CT · axial 182
MRI · T1·T2·DWI
PET · SUV
12,384 feat.
Latent · z
z₁ · 0.412
z₂ · −0.091
z₃ · 0.887
1,024-dim
Molecular
Proteomic · 412
WGS · variants
RNA · GSEA
8,120 markers
Early detect+14 mo
Strat. AUC0.93
Response C-idx0.81
06  / PARTNERSHIPS
Institutional Partnerships.
CLINICAL RESEARCH NETWORK · INDIA

Built with leading institutions.

Active collaborations
01
Council of Scientific & Industrial ResearchNational Institute for Interdisciplinary Science and Technology
Thiruvananthapuram
02
Regional Cancer CentrePremier oncology research & treatment institution
Thiruvananthapuram
03
Malabar Cancer CentreComprehensive cancer care & research
Calicut
07  / WHO IT'S FOR
Three doors into the platform.
INVEST · DEPLOY · RESEARCH
For investors

A platform bet on quantitative oncology.

Series A closed Q1 2026. $42M led by Polaris Bio, joined by Khosla Ventures, GV, and strategic from three top-20 pharma.

  • 4 commercial indications, 2 in late-stage trial
  • 11 hospital systems · 3 pharma co-development deals
  • FDA 510(k) submitted Q4 2025 · CE-MDR Q2 2026
  • 18 publications · 34-institution training corpus
Request investor deck
For hospitals

Your scanners already saw it. Read it.

DICOM-in, JSON-out, on-prem or cloud. Integrates with PACS/RIS in under a week; radiologist workflow-native reports with confidence intervals and audit trail.

  • <90s per study · on-prem GPU or HIPAA cloud
  • Epic · Cerner · Sectra · Philips ISP integrations
  • Structured reports · DICOM-SR · HL7 FHIR
  • HIPAA · SOC 2 Type II · ISO 13485
Schedule a pilot
For researchers

A biomarker engine, not a black box.

Academic license for CanScan features, pre-trained encoders, and the 12,384-feature radiomic panel. Pull imaging embeddings into your own analyses.

  • Academic API · 500k free studies / quarter
  • Pre-trained encoders · ONNX + PyTorch
  • Open feature panel · reproducible pipelines
  • Collaboration desk · 34 active studies
Apply for access
08  / CAREERS
Help us read the invisible.
5 OPEN ROLES · ALL HANDS-ON