NetZeroAICT: Revolutionize CT imaging with AI "Digital Contrast" for a climate-neutral and patient-safe healthcare future

The Challenge

The NetZeroAICT consortium is facing the monumental challenge rooted in the widespread use of Contrast Enhanced Computed Tomography (CECT) scans. While vital for diagnosis, CECTs carried a heavy burden: a significant environmental footprint, including 42,000 tonnes of single-use packaging, 900 tonnes of surgical steel, 90,000 tonnes of plastic tubing, and a staggering 150,000,000 kWh of energy consumption annually. Each scan added approximately 9.2 kg of CO2 to the atmosphere, contributing to the healthcare system's 4.4% share of global CO2 emissions. Compounding this, 200,000 tonnes of iodine contamination entered water systems yearly, a recognized form of "pharmaceutical pollution".

Beyond the environmental toll, CECTs posed real patient risks, from uncomfortable needle insertions to potential kidney failure and severe allergic reactions. The consortium saw an urgent opportunity to harness artificial intelligence to develop "CT Digital Contrast," a revolutionary approach to synthesize contrast digitally from non-contrast scans, thereby avoiding the need for radiocontrast media entirely and paving the way for a healthier planet and safer patients.

The Solution

The NetZeroAICT consortium’s decision to partner with Collective Minds was a strategic masterstroke, building upon the foundational successes of the prior AICT consortium. They understood that advancing medical imaging into a sustainable future required a partner with unparalleled expertise in compliant data infrastructure and a deep understanding of complex medical workflows. Collective Minds emerged as the natural coordinator, bringing a proven track record and an innovative platform.

What truly set Collective Minds apart was their ability to provide a secure, GDPR-compliant cloud platform engineered for clinical and imaging data sharing on a massive scale. They offered a unique solution to the critical challenge of siloed data, providing access to a vast network of 25,000 connected healthcare professionals and specializing in bespoke workflows for academic and industry clinical trials.

The consortium started by leveraging Collective Minds' platform to build the "most comprehensive and diverse CT image repository". The platform’s unique features immediately enhanced their operations

• Zero-Footprint Web Technology: Enabling seamless, scalable data transfer and storage leveraging cloud computing, minimizing local infrastructure needs.
• Subject Pseudonymization: Implementing a cutting-edge, unidirectional hashing system for on-premises de-identification of data, ensuring robust privacy while maintaining full traceability for data controllers.
• Extensibility via API: Providing secure external connectivity that empowers consortium partners to curate, harmonize, and access data for their AI model development and validation efforts.
• Advanced Cohort Management: Offering intuitive user experience (UX) via a web frontend and API, allowing detailed searching, filtering, and online viewing of data. This orchestrates analysis and processing pipelines for different parameterizations with assured reproducibility.
• World-Class Security: Operating within an AWS Frankfurt-hosted Virtual Private Cloud (VPC), with data encrypted both in movement and at rest, and verified by ISO standards, providing regulatory confidence (MDR, FDA).

The Results

The impact of Collective Minds on the NetZeroAICT consortium has been nothing short of transformative! By providing the backbone of the project—a centralized, secure, and highly functional data repository—Collective Minds has dramatically accelerated the consortium’s mission to innovate sustainable healthcare.

The platform has enabled the consortium to:

• Build an Unrivaled Data Repository: Collective Minds is facilitating the creation of the world's most comprehensive and diverse CT image repository, essential for training and validating advanced AI models. This addresses the critical need for high-quality, vast datasets to drive medical AI development.
• Achieve Seamless Connectivity and Dataflow: The platform provides fully operational APIs and robust data transfer capabilities, connecting 9 clinical partners across three continents. This connectivity ensures efficient data ingestion, curation, and access for all consortium members, allowing for dynamic data enrichment and precise segmentation into specific cohorts.
• Ensure Data Trustworthiness and Compliance: Collective Minds’ strict adherence to GDPR, ethical by design principles, and comprehensive security protocols has established a reliable and transparent framework for data handling. This has significantly bolstered public confidence and assured compliance with critical legal and ethical standards.
• Accelerate AI Model Development and Validation: The streamlined data infrastructure has directly empowered Work Package 4 (AI model training and testing) and Work Package 5 (AI model clinical validation). The clinical validation studies, focusing on five critical Indications for Use (IFUs) like visualizing the aorta and pulmonary arteries, are progressing smoothly, leveraging Collective Minds' platform for image presentation and evaluation.
• Hit Key Milestones: The consortium is on track to meet ambitious goals, including transferring 500,000 CT datasets by month 24 (Milestone 3) and ensuring an annual repository uptime of 99.5% (D5.9-D5.12). Key deliverables like the Data Management Plan (D2.1, D2.2) and Legal Framework agreements (D1.1) were produced, and clinical study approvals (D5.13) were secured with impressive efficiency.