Luminescence Image Mapping for Unified Fluorescence Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Luminescence images, particularly fluorescence images, acquired during medical procedures often have varying dynamic ranges, leading to challenges in accurately identifying and quantifying fluorescence agents due to interference from disturbing light, which can result in misclassification of body-part locations and affect surgical, diagnostic, and therapeutic outcomes.
Innovation Solution
A method for processing luminescence images by determining a global range and mapping their values to a common dynamic range, followed by segmenting the images using a unified segmentation threshold, thereby normalizing and segmenting the images for accurate representation and segmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If individual segmentation thresholds are calculated for each fluorescence image based on local statistical distribution, then segmentation can be performed on each image independently, but disturbing light biases the statistical distribution and causes misclassification of body-part locations
Solution Approach 1:
The patent combines multiple fluorescence images into a composite image by accumulating signal values across images. This merging approach increases the signal-to-noise ratio and creates a more robust statistical distribution that is less susceptible to bias from disturbing light in any single image, thereby improving segmentation accuracy while maintaining operational efficiency.
Solution Approach 2:
The patent creates a virtual composite representation (copy) of the fluorescence data across multiple images. By working with this synthesized composite image rather than individual images, the system achieves more reliable segmentation thresholds that reflect the true fluorescence distribution while being resistant to transient disturbances in individual frames.
2Adaptability or versatility
If fluorescence images with different dynamic ranges are displayed individually, then each image can be optimized for its specific range, but accurate identification and quantification of fluorescence agents becomes difficult
Solution Approach 1:
The patent applies logarithmic intensity compression to normalize the dynamic range across multiple fluorescence images with different exposure settings. This transformation creates an equipotential display where images acquired at different times and conditions can be viewed and compared on the same intensity scale, enabling accurate identification and quantification of fluorescence agents regardless of individual image dynamic range variations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy of lesion recognition in surgical procedures, improves diagnostic identification and quantification, and ensures precise delineation of treatment areas, reducing the risk of misclassification and improving patient health outcomes.
Implementation Method 1
Luminescence imaging is based on a luminescence phenomenon, consisting of the emission of light by luminescence substances when subject to any excitation different from heating; particularly, a fluorescence phenomenon occurs in fluorescence substances (called fluorophores), which emit (fluorescence) light when they are illuminated.
Data Source
Figure 1
Figure 2A~2B
Figure 2C~2D
AI summary
A solution for imaging one or more body-parts (109) of a patient (106) in a medical application is proposed. A corresponding method (500) comprises determining (533-536) a mapping function that maps a global range of the values of a plurality of luminescence images of the body-parts (109) to a (common) dynamic range of a displayer (169). The luminescence images are mapped (539-545) by converting their values according to the mapping function. The resulting mapped luminescence images are then displayed (584) together on the displayer (169). In addition or in alternative, the method (500) comprises providing (557) at least one (common) segmentation threshold for starting images equal to the luminescence images or to the mapped luminescence images. The starting images are segmented (560-569) each according to a comparison of the values of the starting image with the segmentation threshold. A computer program (400) and a corresponding computer program product for implementing the method (500) are also proposed. Moreover, a corresponding computing device (121) and an imaging system (100) comprising it are proposed. A surgical method, a diagnostic method and a therapeutic method based on the same method (500) are further proposed.