System of Computational Intelligence Algorithms for Cardiological Diagnostic Support of Predictive Analysis of 4 Short-Plane Ultrasound Projections

Project Information

LPTW Sp. z o.o. from Rzeszów is implementing a project entitled “Development of startups through the ‘Deeptech Accelerator’ – an acceleration program run by the Podkarpackie Innovation Center”, under the European Funds for a Modern Economy 2021–2027, Priority II – Innovation-friendly environment, Measure 2.28 Startup Booster Poland – Grant Agreement No. 8/I/FENG/PCI/2025.

Project tasks:

• Development of a new module enabling the analysis of 4 new projections and providing feedback with at least 90% precision regarding the suspicion of Congenital Heart Defect (CHD) in fetal screening examinations, under conditions similar to clinical practice.
• Development of a technically complete module ensuring medical data loading, preprocessing, and a repository of prepared and anonymized data. Preparation of legal documentation.
• Design of solution architectures for detecting congenital heart defects based on ultrasound projections: 4CH, LVOT, RVOT, and 3VT.
• Validation of the solution with the Technology Recipient under near-real clinical conditions. Full integration of the modular solution with the Ultra Pregna Scan system with defined parameters – a solution ready for implementation.
• Expansion of the potential user group of the system and facilitation of training by reducing barriers related to deployment.

Target group:

• Hospitals and clinics, private medical practices / individual medical practitioners.
• Medical device manufacturers.
• Foundations and non-profit organizations supporting CHD diagnostics.
• Individual clients, parents/patients – indirectly influencing the market, seeking centers or doctors using modern technology.

Project goals:

• Implementation of the new KARDIODIAGNOSTICAI module into the Ultra Pregna Scan (UPS) system, utilizing four fetal heart projections (4CH, LVOT, RVOT, 3VT) to improve effectiveness and simplify the detection process of congenital heart defects (CHD) in fetuses.
• Increasing the usability of the UPS system: simplifying the acquisition of fetal heart ultrasound recordings so that it can also be performed by non-expert clinicians.
• Improving diagnostic performance: achieving high diagnostic accuracy comparable to currently used single recording methods, while simplifying acquisition.
• Reducing errors and increasing AI system robustness against data drift: minimizing mistakes resulting from improper or manually complex ultrasound acquisition.
• Detection of subtle structural changes: enabling more effective identification of even minor heart abnormalities that may indicate CHD.

Effects and outcomes:

• Implementation of the KARDIODIAGNOSTICAI module into the Ultra Pregna Scan system, expanding its functionality in fetal heart ultrasound analysis.
• Simplification of ultrasound acquisition – examinations can also be performed by physicians with limited clinical experience.
• Maintenance of high diagnostic accuracy and reduction of errors caused by improper recording, thanks to automatic projection analysis by the module.
• Increased resilience of the AI system to data drift, improving long-term diagnostic efficiency.
• Greater accessibility of the system for non-expert physicians, increasing the business potential of the solution.
• Improved precision in detecting congenital heart defects in fetuses.

EU project funding: PLN 400,000
Total project value: PLN 436,800
Eligible expenditure amount: PLN 400,000