Liver And Spleen Image Segmentation for Disease Severity Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for classifying liver disease severity from non-invasive radiographic images, such as LiverMultiScan® and FibroScan®, face challenges including lengthy protocols, unreliable readings, variable performance, and inability to analyze historical data, particularly for patients with obesity or ascites, and lack reproducibility and accuracy in fibrosis quantification.
Innovation Solution
A method using radiographic images, pre-processing techniques like segmentation and normalization, combined with machine learning models, to classify liver disease severity by extracting parameters from liver and spleen images, and optionally incorporating patient clinical information, to enhance accuracy and simplify radiological acquisition protocols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MRI is used to measure liver for fibrosis classification, then measurement precision is improved, but device complexity and acquisition time increase
Solution Approach 1:
The liver is segmented into multiple regions of interest (ROIs) within the MRI images. Multiple measurements are taken from different liver segments and combined to improve overall measurement precision and reliability of fibrosis classification
Solution Approach 2:
Standardized pre-processing steps including noise filtering, intensity normalization, and quality assessment are performed on MRI images before measurement. This preliminary action ensures consistent data quality and reduces variability in subsequent fibrosis severity classification
2Measurement precision
If MRI is used for liver disease classification, then measurement precision is improved, but loss of time increases due to lengthy protocol
Solution Approach 1:
The invention extracts and utilizes only the essential MRI sequences and parameters needed for fibrosis classification, eliminating unnecessary scan sequences. This extraction approach maintains measurement precision while significantly reducing total acquisition time
Solution Approach 2:
The method uses a partial set of MRI sequences (focusing on T1-weighted and T2-weighted images with specific parameters) rather than complete liver imaging protocols. This partial action provides sufficient data for accurate fibrosis classification without requiring exhaustive scanning
3Measurement precision
If MRI is used for liver disease classification, then measurement precision is improved, but productivity decreases due to additional resources required
Solution Approach 1:
The MRI protocol is designed to serve multiple functions: it provides diagnostic images for radiologists, enables automated fibrosis classification through extracted parameters, and generates quantitative data for longitudinal monitoring. This multi-functionality increases productivity by eliminating the need for separate dedicated fibrosis assessment scans
4Ease of operation
If FibroScan is used for fibrosis quantification, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The invention uses MRI images as an intermediary that can be processed automatically through standardized pre-processing and measurement pipelines. This intermediary approach maintains ease of operation while achieving superior measurement precision compared to direct ultrasound-based methods like FibroScan
5Measurement precision
If LiverMultiScan protocol is used, then measurement precision is improved for some parameters, but reliability deteriorates for patients with obesity or ascites
Solution Approach 1:
The measurement approach focuses on selecting specific liver regions (segments) that are optimally visualized in each patient's MRI images. By adapting ROI selection to local image quality and patient anatomy, the method maintains reliability even in challenging cases with obesity or ascites
Solution Approach 2:
The invention adjusts measurement parameters and selection criteria based on image quality assessment and patient characteristics. This dynamic parameter adaptation ensures reliable fibrosis classification across diverse patient populations including those with obesity or ascites
6Ease of manufacture
If LiverMultiScan is calibrated on morphological information, then ease of manufacture is improved, but measurement precision deteriorates for fibrosis assessment
Solution Approach 1:
The fibrosis assessment combines multiple types of information: morphological features from liver shape and structure, signal intensity characteristics from T1 and T2-weighted images, and texture features from image analysis. This composite approach achieves high fibrosis measurement precision while remaining implementable with standard MRI sequences
Data Source
AI summary
A method for performing classification of a severity of at least one liver disease from non-invasive radiographic images is disclosed. The method comprises: providing radiographic images of slices of an abdomen of a patient; pre-processing said radiographic images by: segmenting a liver and a spleen, thus achieving a spleen binary mask and a liver binary mask per slice, and normalizing said images with each other, thus achieving normalized radiographic images per slice; for each slice, from the liver binary mask and said normalized radiographic images, extracting a liver parameter; from at least one spleen binary mask, extracting a spleen parameter; and classifying, in function of both parameters and by help of a trained Machine Learning model, the severity of the at least one liver disease between one among a group of liver disease at early stage and a group of liver disease at advanced stage.