Hybrid QUS Radiomics: A Multimodal-Integrated Quantitative Ultrasound Radiomics for Assessing Ambulatory Function in...

Area of Science:

• Medical Imaging
• Biophysics
• Neurology

Background:

• Duchenne muscular dystrophy (DMD) is a neuromuscular disorder impacting ambulatory function.
• Quantitative ultrasound (QUS) imaging with envelope statistics is effective for DMD diagnosis.
• Radiomics extracts detailed features from QUS images.

Purpose of the Study:

• To propose and evaluate a hybrid QUS radiomics approach for characterizing DMD.
• To assess the value of integrating features from multiple QUS imaging models.

Main Methods:

• Eighty-five patients with DMD underwent gastrocnemius ultrasound using Nakagami, homodyned K (HK), and information entropy imaging.
• Hybrid QUS radiomics extracted, selected, and integrated features for ambulatory function classification via support vector machine.
• Performance was evaluated using nested five-fold cross-validation and averaged area under the receiver operating characteristic curve (AUROC).

Main Results:

• Radiomics improved AUROC for B-scan, Nakagami, HK, and entropy imaging (0.790–0.911).
• The hybrid QUS radiomics using HK and entropy images achieved a superior average AUROC of 0.971 for DMD ambulatory function diagnosis.
• This represents a significant improvement over conventional radiomics analysis (p < 0.001).

Conclusions:

• The proposed hybrid QUS radiomics effectively enhances DMD assessment performance.
• This approach integrates microstructure-related backscattering information from various QUS envelope statistics models.
• The findings highlight the potential of hybrid QUS radiomics for improved DMD diagnosis.