Advancing liver disease assessment: a breakthrough in automatic liver segments and quantification using Convolutional Neural Networks (CNN)

Blog
Tags: AIHealth TechLiver diseasePrecision Medicine

Liver diseases, such as non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH), pose significant challenges in diagnosis and monitoring. Evaluating liver deposits, including fat and iron, is crucial in assessing these chronic diffuse liver diseases. 

Traditional methods of liver segmentation are often limited by high computational costs, extended processing times, and difficulties in adapting to different liver anatomies. This scientific paper introduces a novel methodology utilizing CNN for automatic liver segments MRI analysis and quantification.

CNN-based model for liver segmentation and automatic quantification in MRI 

The proposed CNN-based model employs an encoder-decoder structure with four convolutional blocks on each branch. This structure is optimized with deep supervision and normalization techniques, ensuring precise boundary detection and enhanced generalization to diverse liver segments radiology datasets. 

Key features of the CNN-based liver segmentation model:

  • Data augmentation and cross-validation: the training process includes a robust 5-fold cross-validation strategy to evaluate the model’s performance and adaptability.
  • Performance metrics: achieving a high median Dice coefficient of 94% and a false discovery rate of only 4%, the model demonstrates exceptional accuracy in automatic liver segmentation tasks.
  • Quantification of biomarkers: the CNN model effectively quantifies proton density fat fraction (PDFF) and R2* relaxation rates; essential biomarkers for assessing fat and iron deposits in liver segments MRI.

The study further examines the heterogeneity of fat and iron distribution in the liver and confirms the model’s ability to generalize to cases with various distributions. External validation using an independent dataset from different centers and scanners demonstrates the reproducibility of the model across various imaging settings.

Improving diagnosis through a reliable and efficient solution for liver and fat quantification in MR images

Accurate identification of liver segments is pivotal in the diagnosis and monitoring of chronic diffuse liver diseases, providing additional information to radiologists and reducing the limitations associated with manual segmentation methods.

The breakthrough methodology presented in this scientific paper demonstrates the effectiveness and generalizability of a CNN-based model for automatic liver segmentation and quantification. By accurately capturing the liver parenchyma, this approach enables precise assessment of fat and iron deposits, essential for diagnosing and monitoring chronic diffuse liver diseases.

The CNN-based model offers a fast and automatic procedure for MR virtual biopsy, significantly enhancing the clinical evaluation of liver diseases and opening new avenues for research and treatment advancements in this field. Automated solutions like this CNN-based model significantly improve precision, making liver segments radiology assessments faster, more reliable, and highly reproducible.

Liver segments: Figure Image Advancing liver disease
Figure 1. Segmentations were obtained on three different test cases at different anatomical levels. In yellow the CNN predicted segmentation mask is shown and in red the contour of the ground truth is represented. a Results in a study from one of the centers used for the model development. b Results in a study from the other center used for the model development. c Performance on a study which belongs to a patient with partial hepatectomy

Referece

Precise whole liver automatic segmentation and quantification of PDFF and R2* on MR images. Jimenez-Pastor, A., Alberich-Bayarri, A., Lopez-Gonzalez, R. et al. Eur Radiol 31, 7876–7887 (2021) 

Read more

Quibim Website
Privacy Overview

When you visit any website, it may store or retrieve information on your browser, mostly in the form of cookies. This information might be about you, your preferences or your device and is mainly used to make the site work as you expect it to. The information does not directly identify you but can give you a more personalized web experience. Because we respect your right to privacy, you can choose not to allow some types of cookies. Click on the different category headings to learn more and change our default settings. However, blocking some types of cookies may impact your experience of the site. You can find more information, including a detailed cookie explanation, on our Cookie Policy.