↓ this is what we do

Radiology Intelligence
that actually helps

Nexus is a lightweight observability layer for radiology. It reduces diagnostic misses, standardizes follow-up, and catches guideline conflicts — giving your department the oversight it needs for quality assurance and patient safety.

Validated clinical algorithms + visual workflow editor + integrated viewer + on-site deployment. All in one platform, under your control.

HIPAA ✓GDPR ✓On-Site ✓Air-Gapped ✓ISO 13485 (in progress)

* the brains

AI Intelligence — validated & published

Every number here comes from independent research and peer-reviewed publications. Not marketing claims — real clinical evidence.

Lung Oncology Module

✓ Validated

Comprehensive lung nodule analysis with externally validated classifiers and published research results. Covers primary lung tumors and metastatic nodules.

1
Detect
All nodule types: solid, sub-solid, GGO, masses
2
Segment
3D volumetric segmentation with mask overlays
3
Classify
Morphology, composition, margins, shape
4
Score
5-category malignancy scoring with confidence levels
5
Diagnose
Differential: hamartoma, ILN, tuberculoma
6
Track
Longitudinal matching & growth analysis across timepoints

Beyond Chest CT

In Development

Expanding AI capabilities across anatomical regions and imaging modalities.

  • Multi-region lymph node detection (axillary, mediastinal, abdominal, inguinal)
  • Liver & kidney lesion detection and segmentation
  • PET-CT integration with metabolic measurements (SUV-max, SUV-mean)
  • 117 organ segmentations with automated clinical measurements
  • Study routing infrastructure (automatic phase/reconstruction detection)

Clinical AI Agents

In Development

Automated document processing and clinical decision support, running on local LLMs for privacy. Transparent reasoning, not a black box.

  • Document intelligence: scanned PDF to structured FHIR JSON
  • Medical coding: automatic LOINC, SNOMED, ICD-10 assignment
  • Guideline-based patient routing with analysis of existing reports for protocol compliance
  • Transparent reasoning grounded in formalized clinical guidelines
  • Runs on local LLMs — no external API calls, privacy-preserving

* the toolbox

More than algorithms — a complete platform

Nexus is built for real clinical deployment, integration, and governance. Here's what's under the hood:

Workflow Editor

(drag & drop!)

Visual editor for composing AI algorithm pipelines. Build complex processing workflows as directed acyclic graphs (DAGs) without writing a line of code.

2D/3D Medical Viewer

(built right in)

Integrated volume rendering with segmentation overlays, mask toggle, and opacity controls. Review AI findings directly in the platform.

PACS Integration

(plug & play)

Connect to existing infrastructure via DICOM Query/Retrieve (C-FIND, C-MOVE). Seamless integration with your current radiology systems.

On-Site Deployment

(your data stays yours)

Fully containerized and air-gapped capable. Zero external dependencies. Patient data never leaves your facility.

Organization Management

(multi-site ready)

Support for multiple organizations and sites with role-based access control, department-level isolation, and centralized administration.

Audit & Compliance

(every action tracked)

Immutable audit logging with full action history. HIPAA-compliant infrastructure designed for regulatory requirements from the ground up.


* proof it works

Research that speaks for itself

Published results from independent external validations and prospective clinical studies. Not our own benchmarks — outside researchers tested it.

AUC 0.83

Hamartoma classifier

sens. 0.73, spec. 0.93, acc. 0.83

91.7%

ILN sensitivity

97.9% specificity, 91.7% accuracy

+8.8%

Accuracy gain

Malignancy Index (p=0.0003)

+13.4%

Malignant nodules

Classification improvement

ECR 2026

Poster accepted

ILN external validation

Publications & Conferences (the paper trail)

2025
2025
2025
Journal submission ILN classifier validation, submitted to Vestnik Rentgenologii i Radiologii
2026
ECR 2026 ILN classifier external validation poster (Vienna, March 2026)

* who's it for?

Built for your entire team

Nexus serves every role in the radiology workflow — from clinical practice to research and IT deployment.

Radiologists

Thoracic and oncologic radiologists who need AI-assisted detection and diagnosis with transparent, explainable results.

Key capabilities:

  • Review automated findings with segmentation overlays
  • Malignancy scoring with differential diagnosis (hamartoma, lymph node, TB)
  • Longitudinal nodule tracking with growth analysis and VDT

Department Leadership & CMOs

Leadership requiring governance visibility, quality metrics, and workflow orchestration across the department.

Key capabilities:

  • Concordance registries (algorithm vs. physician decisions)
  • Workflow orchestration and algorithm pipeline management
  • Audit trails for regulatory compliance

IT & Deployment Teams

Technical teams responsible for deploying and maintaining clinical infrastructure within existing hospital systems.

Key capabilities:

  • On-site deployment with Docker (air-gapped capable)
  • PACS integration via DICOM (C-FIND, C-MOVE)
  • Organization management and role-based access

Research & Clinical Teams

Researchers and clinicians who need advanced measurements, radiomic features, and structured data for clinical studies.

Key capabilities:

  • Automated measurements across 117 organ segmentations
  • Radiomic feature extraction (107 features per finding)
  • Clinical study data management and export

See Nexus in action?

Experience the full platform — validated clinical algorithms, visual workflow editor, integrated viewer, and on-site deployment. We’ll walk you through how it fits your department.

No pressure, just a friendly demo ✦