DeepHealth, a subsidiary of RadNet, has announced a partnership with CARPL.ai to develop a new artificial intelligence (AI) control system for clinical AI performance and safety.
CARPL.ai specialises in AI orchestration, helping radiologists access, assess and integrate radiology AI solutions into their workflows.
Through this collaboration, DeepHealth will leverage CARPL.ai’s technology to create an AI control system that can be commercialised. The system will optimise imaging AI performance for better clinical outcomes and operational efficiency.
It will also accelerate AI adoption in radiology as AI monitoring is essential for ensuring accurate, reliable, and unbiased performance, CARPL.ai said.
DeepHealth and CARPL.ai will develop a closed-loop AI feedback system. This system will continuously monitor AI model accuracy and relevance in clinical settings.
In addition, it will automate the tracking of performance and safety metrics, including specificity, sensitivity, and data/model drift.
DeepHealth chief operating and technology officer Sham Sokka said: “While the current landscape is marked by an overwhelming array of AI-enabled point solutions, the future involves running multiple AI models, even for a single use case.
“DeepHealth’s partnership with CARPL.ai addresses this very need by creating a unique environment to dynamically run a combination of models and monitor performance and then continuously optimise the best models for specific tasks.”
The partnership will combine CARPL.ai’s AI marketplace and orchestration platform with DeepHealth’s cloud-native operating system, DeepHealth OS.
CARPL.ai’s platform simplifies the selection, implementation, and monitoring of third-party US Food and Drug Administration (FDA)-cleared AI models.
DeepHealth OS unifies data across clinical and operational workflows.
Together, these platforms will be integrated to continuously monitor real-world workflows, enabling radiologists to access safe, high-performing AI interpretation tools seamlessly within their workflows.
CARPL.ai CEO Vidur Mahajan said: “This new AI infrastructure is set to fundamentally redefine radiology by making AI an integral component of the system.
“Monitoring AI performance is essential to ensure the reliability and accuracy of AI applications over time, and our technology enables real-time performance monitoring of both their accuracy and consistency for safe and effective use of AI in clinical practice.”
In a separate development, Siemens Healthineers and DeepHealth enter a partnership to enhance ultrasound operations using SmartTechnology.
This will integrate AI-driven health informatics with Siemens’ Acuson Sequoia ultrasound system for improved efficiency and patient experience in outpatient care.
Recently, GE HealthCare partnered with DeepHealth to advance AI integration in medical imaging.
