Periodic Reporting for period 2 - LVOCheck (Diagnosis of large vessel occlusion (LVO) in acute stroke for improving access to treatment)
Periodo di rendicontazione: 2024-05-01 al 2025-04-30
Stroke is the second leading cause of death and a major source of long-term disability globally. Among stroke types, Large Vessel Occlusion (LVO) is the most severe and time-sensitive. Timely intervention, particularly mechanical thrombectomy, can greatly improve outcomes—but only if patients are rapidly and correctly identified.
Existing prehospital triage methods, typically clinical scales, lack sufficient accuracy, often leading to under- or over-triage, treatment delays, and inefficient use of stroke centers. While Mobile Stroke Units (MSUs) are effective, they remain expensive and largely unavailable, especially in rural or underserved regions.
The LVOCheck project addresses this gap with a portable, AI-supported diagnostic tool designed for ambulance or emergency use. It identifies suspected LVO stroke patients in under 15 minutes by combining three blood-based biomarkers (NT-proBNP, D-Dimer, H-FABP) with a machine learning algorithm accessible via a mobile app, delivering real-time decision support to emergency responders.
Key achievements:
• Developed a point-of-care lateral flow assay for quantifying the biomarker panel
• Created an AI-based clinical decision support (CDS) app integrating biomarkers and clinical indicators
• Conducted two clinical studies (Spain and Germany) with over 800 patients, validating performance in real-world prehospital settings
• Achieved 64–74% sensitivity at 93% specificity, significantly outperforming current clinical triage scores
• Initiated FDA Breakthrough Device Designation submission and advanced CE marking preparation under EU IVDR
• Implemented a QMS compliant with ISO 13485:2016, supporting regulatory scalability
These results confirm LVOCheck’s value as a disruptive diagnostic solution for stroke triage. The system is intuitive, scalable, and fits seamlessly into existing emergency workflows, enabling faster decisions, improved outcomes, and reduced system costs.
Future potential:
LVOCheck can transform stroke triage worldwide, particularly in areas lacking imaging or stroke expertise. By improving access to thrombectomy, it may prevent irreversible brain damage and reduce long-term disability. Health economic modelling suggests meaningful cost savings per patient by streamlining care pathways and avoiding unnecessary transfers.
Moreover, LVOCheck promotes equity in stroke care, bringing cutting-edge diagnostics to underserved regions. Its portability makes it suitable for ambulances, remote clinics, and low-resource settings. The platform also holds promise for expansion into other emergency or cardiovascular indications, leveraging its biomarker-AI synergy.
Conclusion:
LVOCheck represents a major innovation in stroke diagnostics. With strong validation, regulatory momentum, and growing interest from investors and clinical partners, ABCDx is positioned to scale deployment across Europe, North America, and the Gulf region. Support from the EIC Accelerator has been instrumental in translating a bold scientific concept into a market-ready solution with global impact.
Existing prehospital triage methods, typically clinical scales, lack sufficient accuracy, often leading to under- or over-triage, treatment delays, and inefficient use of stroke centers. While Mobile Stroke Units (MSUs) are effective, they remain expensive and largely unavailable, especially in rural or underserved regions.
The LVOCheck project addresses this gap with a portable, AI-supported diagnostic tool designed for ambulance or emergency use. It identifies suspected LVO stroke patients in under 15 minutes by combining three blood-based biomarkers (NT-proBNP, D-Dimer, H-FABP) with a machine learning algorithm accessible via a mobile app, delivering real-time decision support to emergency responders.
Key achievements:
• Developed a point-of-care lateral flow assay for quantifying the biomarker panel
• Created an AI-based clinical decision support (CDS) app integrating biomarkers and clinical indicators
• Conducted two clinical studies (Spain and Germany) with over 800 patients, validating performance in real-world prehospital settings
• Achieved 64–74% sensitivity at 93% specificity, significantly outperforming current clinical triage scores
• Initiated FDA Breakthrough Device Designation submission and advanced CE marking preparation under EU IVDR
• Implemented a QMS compliant with ISO 13485:2016, supporting regulatory scalability
These results confirm LVOCheck’s value as a disruptive diagnostic solution for stroke triage. The system is intuitive, scalable, and fits seamlessly into existing emergency workflows, enabling faster decisions, improved outcomes, and reduced system costs.
Future potential:
LVOCheck can transform stroke triage worldwide, particularly in areas lacking imaging or stroke expertise. By improving access to thrombectomy, it may prevent irreversible brain damage and reduce long-term disability. Health economic modelling suggests meaningful cost savings per patient by streamlining care pathways and avoiding unnecessary transfers.
Moreover, LVOCheck promotes equity in stroke care, bringing cutting-edge diagnostics to underserved regions. Its portability makes it suitable for ambulances, remote clinics, and low-resource settings. The platform also holds promise for expansion into other emergency or cardiovascular indications, leveraging its biomarker-AI synergy.
Conclusion:
LVOCheck represents a major innovation in stroke diagnostics. With strong validation, regulatory momentum, and growing interest from investors and clinical partners, ABCDx is positioned to scale deployment across Europe, North America, and the Gulf region. Support from the EIC Accelerator has been instrumental in translating a bold scientific concept into a market-ready solution with global impact.
The LVOCheck project, led by ABCDx, S.L. focused on the development and validation of an innovative, AI-powered point-of-care diagnostic tool designed for the early triage of Large Vessel Occlusion (LVO) strokes in prehospital settings. The project combined expertise in assay development, software engineering, clinical validation, and regulatory affairs.
A major achievement was the development of a lateral flow immunoassay (LFA) capable of quantifying three stroke-associated biomarkers: NT-proBNP, D-Dimer, and H-FABP. This assay served as the biochemical foundation of the diagnostic device.
In parallel, ABCDx created a machine learning–based Clinical Decision Support (CDS) system, integrated into a mobile application. This system combines biomarker results with clinical parameters to provide real-time stroke triage support to emergency medical professionals.
Clinical validation was performed through two real-world studies in Spain and Germany, involving 807 plasma samples. The device achieved a sensitivity of 64–74% and a specificity of 93%, outperforming traditional triage methods.
To prepare for market access, ABCDx implemented a Quality Management System (QMS) in compliance with ISO 13485:2016 and FDA regulations. A Breakthrough Device Designation request was submitted to the FDA, and regulatory frameworks for EU IVDR and SFDA (Saudi Arabia) were mapped in preparation for global deployment.
A major achievement was the development of a lateral flow immunoassay (LFA) capable of quantifying three stroke-associated biomarkers: NT-proBNP, D-Dimer, and H-FABP. This assay served as the biochemical foundation of the diagnostic device.
In parallel, ABCDx created a machine learning–based Clinical Decision Support (CDS) system, integrated into a mobile application. This system combines biomarker results with clinical parameters to provide real-time stroke triage support to emergency medical professionals.
Clinical validation was performed through two real-world studies in Spain and Germany, involving 807 plasma samples. The device achieved a sensitivity of 64–74% and a specificity of 93%, outperforming traditional triage methods.
To prepare for market access, ABCDx implemented a Quality Management System (QMS) in compliance with ISO 13485:2016 and FDA regulations. A Breakthrough Device Designation request was submitted to the FDA, and regulatory frameworks for EU IVDR and SFDA (Saudi Arabia) were mapped in preparation for global deployment.
1. Overview of Results
The LVOCheck project has delivered a first-in-class diagnostic platform that enables rapid, biomarker- and AI-based triage of Large Vessel Occlusion (LVO) stroke patients in prehospital settings. Unlike current stroke triage methods that rely on subjective clinical scales with limited predictive power, LVOCheck introduces a fully integrated solution combining:
• A multi-biomarker lateral flow assay (LFA) targeting NT-proBNP, D-Dimer, and H-FABP
• A mobile Clinical Decision Support (CDS) app using machine learning algorithms
• A cloud-based backend for analysis and secure data handling
• Real-world validation across two international clinical cohorts (Spain and Germany)
By delivering a sensitivity of 64–74% at 93% specificity, LVOCheck demonstrates substantial superiority over standard-of-care prehospital tools, such as NIHSS, RACE, or FAST-ED, which typically fail to exceed 50% sensitivity at similar specificity thresholds. This leap in performance enables earlier access to thrombectomy, improves patient outcomes, and reduces unnecessary patient transfers.
2. Key Innovations and Breakthroughs
• First AI-driven stroke triage tool combining biomarkers and clinical parameters
• First use of a three-biomarker panel (NT-proBNP, D-Dimer, H-FABP) for LVO identification in emergency medicine
• First portable LVO diagnostic validated in real-world prehospital settings (ambulances)
• Demonstrated usability and acceptability by non-specialized emergency professionals
• Validated smartphone interface with cross-device compatibility
The LVOCheck project has delivered a first-in-class diagnostic platform that enables rapid, biomarker- and AI-based triage of Large Vessel Occlusion (LVO) stroke patients in prehospital settings. Unlike current stroke triage methods that rely on subjective clinical scales with limited predictive power, LVOCheck introduces a fully integrated solution combining:
• A multi-biomarker lateral flow assay (LFA) targeting NT-proBNP, D-Dimer, and H-FABP
• A mobile Clinical Decision Support (CDS) app using machine learning algorithms
• A cloud-based backend for analysis and secure data handling
• Real-world validation across two international clinical cohorts (Spain and Germany)
By delivering a sensitivity of 64–74% at 93% specificity, LVOCheck demonstrates substantial superiority over standard-of-care prehospital tools, such as NIHSS, RACE, or FAST-ED, which typically fail to exceed 50% sensitivity at similar specificity thresholds. This leap in performance enables earlier access to thrombectomy, improves patient outcomes, and reduces unnecessary patient transfers.
2. Key Innovations and Breakthroughs
• First AI-driven stroke triage tool combining biomarkers and clinical parameters
• First use of a three-biomarker panel (NT-proBNP, D-Dimer, H-FABP) for LVO identification in emergency medicine
• First portable LVO diagnostic validated in real-world prehospital settings (ambulances)
• Demonstrated usability and acceptability by non-specialized emergency professionals
• Validated smartphone interface with cross-device compatibility