Anatomical Vector Segmentation for Faster Medical Image Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for segmenting and labeling medical patient images require significant computational effort due to the need for registering patient images with anatomical atlases.
Innovation Solution
A method involving a tracked imaging device and an anatomical atlas is used to establish a transformation between the atlas space and the patient space, training a learning algorithm with labels and anatomical vectors to enable efficient segmentation and labeling of patient images without repeated atlas registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If patient images are registered with anatomical atlases using traditional methods, then segmentation and labeling accuracy is improved, but computational effort and processing time increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-training a learning algorithm (neural network) on a dataset of anatomical images and their corresponding segmentations before actual use. This training phase is performed in advance, allowing the model to learn anatomical patterns and relationships. During actual segmentation tasks, the pre-trained model can rapidly process new patient images without requiring time-consuming atlas registration, thus reducing processing time while maintaining segmentation accuracy.
Solution Approach 2:
The patent replaces the mechanical system of iterative atlas registration with a learned mapping approach using neural networks. Instead of computationally intensive image registration algorithms that mechanically align images through transformation parameters, the system uses a trained neural network that has learned the transformation patterns during training. This substitution of mechanical registration with a learned model significantly reduces computational effort and processing time for segmentation tasks.
2Measurement precision
If traditional atlas-based segmentation is used, then anatomical accuracy is improved, but device complexity and computational resources required increase
Solution Approach 1:
The patent uses copying by creating a learned representation (model) of anatomical structures from training data. Instead of requiring the full complexity of atlas-based registration systems, the approach copies the essential anatomical patterns and relationships into a neural network model. This copied knowledge allows the system to perform segmentation with high anatomical accuracy while requiring simpler computational infrastructure during inference, as the complex patterns have already been captured during training.
3Measurement precision
If repeated atlas registration is performed for each patient image, then consistent anatomical labeling is improved, but productivity and throughput decrease
Solution Approach 1:
The patent applies preliminary action by pre-training the learning algorithm on a comprehensive dataset that encompasses various anatomical variations and perspectives. This pre-training establishes consistent labeling patterns across diverse images. When processing new patient images, the pre-trained model applies these learned patterns directly without requiring repeated registration steps, thereby maintaining labeling consistency while significantly improving processing throughput and productivity.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Disclosed is a computer-implemented method which encompasses registering a tracked imaging device such as a microscope having a known viewing direction and an atlas to a patient space so that a transformation can be established between the atlas space and the reference system for defining positions in images of an anatomical structure of the patient. Labels are associated with certain constituents of the images and are input into a learning algorithm such as a machine learning algorithm, for example a convolutional neural network, together with the medical images and an anatomical vector and for example also the atlas to train the learning algorithm for automatic segmentation of patient images generated with the tracked imaging device. The trained learning algorithm then allows for efficient segmentation and/or labelling of patient images without having to register the patient images to the atlas each time, thereby saving on computational effort.