Optical Imaging CT Parameter Determination

The patent replaces the x-ray based topogram acquisition system with an optical imaging system. Optical cameras capture images of the patient's body surface, and machine learning algorithms process these optical images to estimate patient body size and determine CT scan parameters. This substitution eliminates ionizing radiation exposure while maintaining the ability to accurately assess patient anatomy for dose calculation.

The patent introduces optical imaging data and machine learning algorithms as intermediaries between the patient and the CT scan parameter determination process. Instead of directly using x-rays to create topograms, the system uses optical cameras to capture surface images, processes them through neural networks to extract body size information, and then uses this information to determine appropriate CT scan parameters. This intermediary approach achieves the same goal without the harmful radiation effects.

The patent performs optical imaging and machine learning processing in advance of the actual CT scan to determine patient body size and establish initial scan parameters. By completing the measurement and parameter determination process before the CT scan begins, the system eliminates the time that would otherwise be spent acquiring topograms and processing them sequentially. The optical imaging captures all necessary information simultaneously, and the neural network processes this data rapidly to provide immediate results for scan parameter determination.

The patent fundamentally changes the measurement parameters from x-ray attenuation measurements (topograms) to optical surface imaging parameters. Optical cameras capture reflectivity and geometric information from the patient's body surface, which are then transformed by machine learning models into body size estimates. This parameter change enables parallel processing of multiple body regions simultaneously and eliminates the sequential acquisition time required for traditional topograms, thereby reducing overall scan time while maintaining measurement accuracy.

The patent replaces direct x-ray attenuation measurements with optical imaging combined with machine learning. Optical cameras capture images of the patient's body surface, and trained neural networks process these images to generate attenuation curves that are equivalent to those obtained from topograms. The machine learning models are trained on large datasets to learn the complex relationships between optical surface properties and underlying tissue attenuation characteristics, enabling accurate attenuation curve generation without ionizing radiation.

The patent creates a computational model (attenuation curves) that copies the essential information normally obtained from x-ray topograms. Instead of directly measuring x-ray attenuation, the system uses optical imaging to capture surface characteristics and then uses machine learning to generate attenuation curves that replicate the information needed for CT dose calculation. This copying approach allows the system to obtain the necessary measurement data through a non-ionizing modality while maintaining the accuracy required for clinical decision-making.