Deep Learning-Augmented Zero Echo Time MRI Increases Diagnostic Confidence for Osseous Assessment in Hand and Foot...

Area of Science:

• Radiology
• Medical Imaging
• Deep Learning Applications

Background:

• Standard MRI protocols for hands and feet can be enhanced with advanced sequences.
• Assessing bone pathology in extremities requires high-resolution imaging techniques.

Purpose of the Study:

• To evaluate the integration of deep learning-augmented zero echo time (ZTE DL) MRI sequences into standard MRI protocols.
• To compare the diagnostic confidence and findings of ZTE DL-augmented MRI with conventional MRI and radiographs for hand and foot pathologies.

Main Methods:

• A retrospective study comparing standard 1.5 T MRI protocols with added ZTE DL sequences for hands and feet.
• Assessment of bone pathologies (sclerosis, osteophytes, joint narrowing, fractures) and indeterminate lesions.
• Diagnostic confidence rated on a 3-point scale; comparisons made using mixed-effects models and Wilcoxon signed-rank tests.

Main Results:

• ZTE DL sequences showed similar findings to conventional MRI but with significantly higher diagnostic confidence for bone changes (p<0.05).
• Compared to radiographs, ZTE DL-augmented MRI demonstrated higher confidence for five of six pathological bone changes (p<0.05).
• Interreader agreement for diagnostic confidence was moderate to substantial (kappa 0.59-0.71).

Conclusions:

• The addition of ZTE DL sequences to standard MRI protocols for hands and feet improves diagnostic confidence.
• ZTE DL-augmented MRI offers comparable findings to conventional MRI and superior confidence over radiographs for bone pathologies.
• Further studies are necessary to validate the diagnostic accuracy of ZTE DL sequences against a reference standard.