VIRTUAL CLINICAL TRIALS IN MEDICAL IMAGING SYSTEM EVALUATION AND OPTIMISATION.

Area of Science:

• Medical Imaging
• Computational Biology
• Radiology

Background:

• Virtual clinical trials (VCTs) offer a computational approach to medical imaging research.
• The OpenVCT framework facilitates VCTs, particularly for breast imaging applications.
• Evaluating VCTs is crucial for their adoption in optimizing medical imaging systems.

Purpose of the Study:

• To assess the OpenVCT framework's utility in virtual clinical trials for digital breast tomosynthesis (DBT).
• To compare virtual and clinical reading studies for lesion detection in mammography and DBT.
• To optimize a novel simultaneous DBT and mechanical imaging (MI) system using VCTs.

Main Methods:

• Utilized the OpenVCT framework for computer simulations of human anatomy and imaging modalities.
• Conducted virtual and clinical reading studies for lesion detection in DBT.
• Expanded the framework to include mechanical imaging (MI) for system optimization.

Main Results:

• Demonstrated close agreement between virtual and clinical studies, confirming VCTs' predictive accuracy for DBT performance.
• Showcased optimization of simultaneous DBT and MI system, particularly improving DBT image quality.
• Identified areas for future development, including more accurate MI acquisition simulation and tumor growth models.

Conclusions:

• Virtual clinical trials, powered by frameworks like OpenVCT, can accurately predict medical imaging system performance.
• The OpenVCT framework is effective for evaluating and optimizing DBT systems.
• VCTs hold significant future potential in medical imaging, AI training, and rare disease research.