In situtumor model for longitudinal in silico imaging trials

Area of Science:

• Computational modeling
• Medical imaging
• Biophysics

Background:

• Breast cancer progression is influenced by the mechanical properties of surrounding tissues.
• Understanding how tumor cells interact with varying tissue elasticity is crucial for accurate modeling.

Purpose of the Study:

• To develop a computational model simulating breast cancer lesion growth based on anatomical structure stiffness.
• To investigate how differential tissue resistance affects tumor morphology.
• To generate realistic digital mammogram (DM) images of simulated lesions for validation.

Main Methods:

• Utilized a computational model incorporating tissue elasticity (ligaments as rigid, fat/glandular tissue as soft).
• Simulated lesion growth within fatty breast models from the Virtual Imaging Clinical Trials for Regulatory Evaluation (VICTRE) pipeline.
• Generated digital mammograms using a simulated mammography system (Siemens Mammomat Inspiration) for visualization and validation.
• Conducted a reader study comparing simulated lesions with real mammograms from the DDSM dataset.

Main Results:

• The model demonstrated that simulated cancer cells preferentially grow in softer tissues, circumventing stiffer structures like ligaments.
• Digital mammogram images effectively illustrated lesion integration with anatomical backgrounds.
• Reader study indicated the simulated lesions' realism when compared to real mammograms.
• Average simulation time was approximately 2.5 hours, varying with the local environment.

Conclusions:

• The developed lesion growth model accurately simulates breast cancer progression influenced by anatomical topography.
• This model enhances the capability for longitudinal in silico trials in breast cancer research.
• The findings support the use of computational models for understanding tumor behavior in diverse tissue environments.