What challenges did NICO.LAB face before AI implementation?

Stroke treatment is extremely time-sensitive; delays result in loss of healthy living years and increased risk of death or disability. Hospitals face challenges processing large CT scan images quickly and securely, with varying infrastructure and broadband speeds. Speed and accuracy in stroke image analysis are critical to improve treatment outcomes and reduce delays.

What was the impact of the AI solution?

StrokeViewer reduces image processing time to under 3 minutes, accelerating time to stroke treatment and improving patient prognosis. Clinical studies show a 29% reduction in delay to treatment and potential to detect 15 extra strokes per 100 patients. NICO.LAB's solution has FDA approval and is deployed in several Dutch hospitals, with plans for expansion in 6 more European countries, the US, and Australia. The cloud-based solution requires no additional hospital hardware, simplifying adoption and scalability.

Source

GCP

NICO.LAB: Revolutionizing Stroke Care with AI on Google Cloud

Use case typeAI development platformUpdated Jun 13, 2026

NICO. LAB, a Dutch health technology company, developed StrokeViewer, a cloud-based AI solution to speed up stroke diagnosis and treatment. The solution analyzes complex CT scan images in under 3 minutes using TensorFlow ML models running on Google Cloud's Compute Engine with custom GPUs for high performance. It provides highly accurate, clinically validated, and FDA-certified algorithms to identify stroke biomarkers and reduce time to treatment by 29%, improving patient outcomes. The system uses Google Kubernetes Engine for scalable image data pipeline, Cloud Storage for raw data, Cloud Pub/Sub for processing triggers, and the Cloud Healthcare API for secure DICOM image handling. Deployed in multiple hospitals in the Netherlands, it enables easy integration with hospital infrastructure without additional hardware, and is expanding globally.

Organization: NICO.LAB
Industry: Healthcare
Location: Netherlands

Reported outcomes

−29%

quantified impactTime & speed

3 minutestime

Strategic outcomes

Speed & agilityAccelerated stroke diagnosis and treatmentCustomer experience & trustImproved patient outcomesRisk & complianceFDA-approved compliant imaging workflowMarket & geographic expansionExpanded into international healthcare markets

Catalog median for time & speed deployments: −60% across 727 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1Medical Image Analysis
2AI-based Diagnosis
3Healthcare AI

Stroke treatment is extremely time-sensitive; delays result in loss of healthy living years and increased risk of death or disability.
Hospitals face challenges processing large CT scan images quickly and securely, with varying infrastructure and broadband speeds.
Speed and accuracy in stroke image analysis are critical to improve treatment outcomes and reduce delays.

NICO.LAB built StrokeViewer on Google Cloud using TensorFlow for machine learning and Compute Engine VMs with custom GPUs for fast image processing within under 3 minutes.
The image pipeline is managed by Google Kubernetes Engine scaling with demand, storing data in Cloud Storage, and triggering analytics via Pub/Sub.
Secure upload of large CT scan images is ensured using encrypted DICOM protocols and Google Cloud VPN for legacy hospital systems.
Cloud Healthcare API supports DICOM for healthcare compliance and managing imaging data securely in the cloud.

Technologies

TensorFlow Compute Engine Google Kubernetes Engine Cloud Storage Cloud Pub/Sub Cloud Healthcare API

StrokeViewer reduces image processing time to under 3 minutes, accelerating time to stroke treatment and improving patient prognosis.
Clinical studies show a 29% reduction in delay to treatment and potential to detect 15 extra strokes per 100 patients.
NICO.LAB's solution has FDA approval and is deployed in several Dutch hospitals, with plans for expansion in 6 more European countries, the US, and Australia.
The cloud-based solution requires no additional hospital hardware, simplifying adoption and scalability.

Architecture

The architecture uses TensorFlow for ML model building, Compute Engine VMs with custom GPUs for rapid image processing, Google Kubernetes Engine for scalable pipeline management, Cloud Storage for raw image and metadata storage, Cloud Pub/Sub for asynchronous processing triggers, and Cloud Healthcare API to manage DICOM medical image standards securely. Secure image upload is supported by encrypted DICOM protocols and optional Google Cloud VPN.

Sources & evidence1

Groundedness: 5/5

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence