Simulating Light Propagation in a Reconstructed Model from Breast DICOM MRI Images for Developing Optical-Based Diagnosis Modality

Mammography is the most popular and effective way to detect early, non-palpable breast cancer. However, mammography only shows breast structure and cannot tell much about the physiological state of the breast. Clinical examination or self-testing methods have false-positive rates and unwanted side effects, so a new diagnostic method with imaging images is needed to facilitate diagnosis. Optical imaging modalities like diffuse optical tomography (DOT) or transillumination imaging have been proposed with the desire to develop a patient-friendly, low-cost, non-invasive technique that does not use non-ionizing radiation. These novel imaging modalities can complement traditional methods as an additional screening tool. This study investigates a three-dimensional (3D) breast reconstructed model from the DICOM Magnetic resonance imaging (MRI) dataset for developing optical-based diagnosis applications. The 3D model is segmented and reconstructed using The Mimicss Innovation Suite software. Then, the 3D breast model’s light propagation is simulated using Monte Carlo-based simulation software. The model is investigated with different scenarios of light sources and detectors. As a result, the absorption distribution, fluence, reflectance, and transmittance show the feasibility of developing a breast optical imaging system. A result of an abnormal object added into the 3D model proves the feasibility of obtaining the abnormal object image. With the view toward developing a breast optical imaging system, the light propagation simulations are conducted by modeling the light propagated in a 3D model with different light sources and detector scenarios. The result validated and confirmed the feasibility of detecting abnormal tissue in breast images by simulation.