Sign in Subscribe
Machine learning

Automated Segmentation of Lumbar Paraspinal Muscles Shows Promise in Chronic Low Back Pain Management

Strategic Insights

— 1 min read

Chronic low back pain is a global health issue with significant socioeconomic burdens, affecting millions of people worldwide. Researchers at Griffith University have made a breakthrough in developing an automated method for segmenting and analyzing lumbar paraspinal muscles (LPM) using artificial intelligence. This innovative approach has the potential to improve the diagnosis and treatment of chronic low back pain.

Key Takeaways:

  • The study used a deep learning method to train a model on 1,302 MRIs from 641 participants across five centers.
  • The model achieved high accuracy in segmenting LPM and demonstrated statistical equivalence to manual measurements of muscle volume and fatty infiltration ratio.
  • The proposed automated method accurately segmented LPM across multisequence, multicenter MRIs.
  • The study's results have significant implications for the diagnosis and treatment of chronic low back pain.
  • The researchers plan to further evaluate the model's performance on larger datasets and in clinical settings.
  • The study has been peer-reviewed and published in Radiology Artificial Intelligence.
  • Additional authors for this research include Zhongyi Zhang, Enrico De Martino, Janet Millner, and Gervase Tuxworth.

Statistics:

  • Total MRIs used in the study: 1,302
  • Number of participants: 641
  • Number of centers: 5
  • Model's accuracy in segmenting LPM: 0.93 to 0.97 (Dice similarity coefficient)
  • Model's accuracy in measuring muscle volume: 0.92 (statistical equivalence between automated and manual measurements)
  • Model's accuracy in measuring fatty infiltration ratio: 0.92 (statistical equivalence between automated and manual measurements)

Sources:

  • Multicenter Validation of Automated Segmentation and Composition Analysis of Lumbar Paraspinal Muscles Using Multisequence MRI. Radiology Artificial Intelligence, 2025.
  • NewsRx. New Artificial Intelligence Data Have Been Reported by Researchers at Griffith University (Multicenter Validation of Automated Segmentation and Composition Analysis of Lumbar Paraspinal Muscles Using Multisequence MRI). Robotics & Machine Learning. September 15, 2025; p 293.

Read more