A Deep-Learning Approach to Spleen Volume Estimation in Patients with Gaucher Disease
Ido Azuri1, Ameer Wattad2, Keren Peri-Hanania3
1Bioinformatics Unit, Department of Life Sciences Core Facilities, Weizmann Institute of Science, Rehovot 7610001, Israel.
Journal of Clinical Medicine
|August 26, 2023
Summary
Deep learning significantly improves spleen volume estimation in Gaucher disease (GD) patients. This advanced method offers greater accuracy than traditional formulas, aiding in better disease monitoring and treatment assessment.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Gaucher Disease Research
Background:
- Hepatosplenomegaly is a common sign of Gaucher disease (GD).
- Accurate liver and spleen volume measurement using MRI is vital for GD assessment and treatment monitoring.
- Current spleen volume estimation formulas in clinical practice have significant inaccuracies.
Purpose of the Study:
- To develop and validate a deep-learning approach for precise spleen segmentation and volume calculation in Gaucher disease patients.
- To compare the accuracy of deep learning-based spleen volume estimation against traditional methods.
Main Methods:
- Utilized deep-learning techniques for spleen segmentation on MRI scans.
- Calculated spleen volumes using the developed deep-learning model.
- Compared results with a conventional formula-based method on a cohort of 20 GD patients.
Main Results:
- The deep-learning approach achieved a mean error of 3.6 ± 2.7% for spleen volume estimation.
- The traditional formula approach resulted in a significantly higher mean error of 13.9 ± 9.6%.
- Deep learning demonstrated superior accuracy in calculating spleen volume.
Conclusions:
- Deep learning offers a more accurate and reliable method for spleen volume estimation in Gaucher disease.
- Integrating deep learning into clinical practice can enhance diagnostic capabilities and treatment monitoring for GD.
- This approach holds potential for improved patient outcomes in Gaucher disease management.
