Non-Contrast CT Vessel Segmentation Without Contrast Agents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Patients are exposed to unnecessary x-ray radiation and potential health risks due to the need for both non-contrast and contrast CT scans, and the administration of contrast agents causes discomfort and health issues, particularly in vulnerable populations.
Innovation Solution
Utilizing non-contrast CT images to train machine learning algorithms for image segmentation and classification, enabling the identification of blood vessel features without the need for contrast agents by generating pseudo-contrast CT images or directly identifying structural features in non-contrast images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If both non-contrast and contrast CT scans are performed to ensure diagnostic accuracy, then diagnostic reliability is improved, but patient exposure to x-ray radiation increases
Solution Approach 1:
The system performs image segmentation and feature extraction on non-contrast CT images before contrast administration. By pre-identifying blood vessels and anatomical structures using machine learning algorithms trained on non-contrast images, the system eliminates the need for a subsequent contrast CT scan, thereby reducing cumulative radiation exposure while maintaining diagnostic capability
Solution Approach 2:
The system creates virtual contrast-enhanced images by applying segmentation masks and synthetic contrast information to non-contrast CT images. This virtual copying of contrast enhancement effects through computational methods allows diagnostic evaluation without actual contrast agent administration and without requiring a separate contrast CT scan
2Measurement precision
If contrast agents are administered to improve blood vessel visualization, then image quality is improved, but patient comfort and health safety deteriorate
Solution Approach 1:
The system enables non-contrast CT images to serve themselves by using machine learning algorithms to automatically identify and segment blood vessels without external contrast agents. The trained models detect vascular structures inherent in non-contrast images, making the imaging process self-sufficient and eliminating dependence on contrast agents
Solution Approach 2:
The system replaces the chemical mechanism of contrast agents with a computational mechanism. Instead of using iodine-based contrast materials to enhance x-ray attenuation differences, the system uses trained neural networks and image processing algorithms to identify and highlight blood vessel structures through pattern recognition in non-contrast images
3Reliability
If repeated CT scans are performed for monitoring patient conditions, then diagnostic accuracy is maintained, but cumulative radiation exposure increases
Solution Approach 1:
The system performs comprehensive image segmentation and feature extraction on initial non-contrast CT images before any contrast administration. By establishing baseline vascular anatomy and pathology through AI analysis of non-contrast images, the system enables longitudinal monitoring through comparison of subsequent non-contrast scans without accumulating radiation from repeated contrast studies
Data Source
AI summary
Methods for training an algorithm to identify structural anatomical features, for example of a blood vessel, in a non-contrast computed tomography (NCT) image are described herein. The algorithm may comprise an image segmentation algorithm, a random forest classifier, or a generative adversarial network in examples described herein. In one embodiment, a method comprises receiving a labelled training set for a machine learning image segmentation algorithm. The labelled training set comprising a plurality of NCT images, each NCT image of the plurality of NCT images showing a targeted region of a subject, the targeted region including at least one blood vessel. The labelled training set further comprises a corresponding plurality of segmentation masks, each segmentation mask labelling at least one structural feature of a blood vessel in a corresponding NCT image of the plurality of NCT images. The method further comprises training a machine learning image segmentation algorithm, using the plurality of NCT images and the corresponding plurality of segmentation masks, to learn features of the NCT images that correspond to structural features of the blood vessels labelled in the segmentation masks, and output a trained image segmentation model. The method further comprises outputting the trained image segmentation model usable for identifying structural features of a blood vessel in an NCT image. Further methods are described herein for identifying anatomical features from an NCT image, and for establishing training sets. Computing apparatuses and computer readable media are also described herein.