Model checking for trigger loss detection during Doppler ultrasound-guided fetal cardiovascular MRI

Area of Science:

• Medical Imaging
• Computer Science
• Cardiology

Background:

• Ultrasound (US) is standard for prenatal fetal heart diagnosis.
• Cardiovascular magnetic resonance imaging (CMRI) offers complementary insights.
• Current CMRI requires trigger-based methods for necessary temporal and spatial resolution.

Purpose of the Study:

• To address trigger loss in Doppler US (DUS)-based triggering for CMRI, which hinders image acquisition.
• To develop a method for accurately deciding when to inject triggers, minimizing reliance on historical data.
• To enhance the reliability of DUS triggering for high-field CMRI.

Main Methods:

• Utilizing model checking, a computer science technique, to formally verify system requirements.
• Simultaneously modeling fetal heart and respiratory motion.
• Employing physiological parameter bounds to detect respiration-induced signal changes and inform trigger injection decisions.

Main Results:

• Reduced falsely injected triggers by 94% in real-world scenarios (from >87% to <5%).
• Achieved a reduction to below 0.2% for motion subsets suitable for CMRI.
• Validated the method's robustness across various motions, ranges, and heartbeat traces using robotic simulations.

Conclusions:

• Accurate trigger injection is crucial for DUS-guided fetal CMRI.
• The proposed model checking method substantially decreases erroneous trigger injections.
• This advancement is a key step towards achieving rapid and artifact-free fetal CMRI.