Advancements in Frank's sign Identification using deep learning on 3D brain MRI

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Radiology

Background:

• Frank's sign (FS) is a clinical marker linked to aging and various health conditions.
• Current methods for identifying Frank's sign lack standardization, hindering consistent clinical application.

Purpose of the Study:

• To develop and validate a deep learning model for automated detection and segmentation of Frank's sign in 3D brain MRI scans.

Main Methods:

• Four deep learning architectures were evaluated for Frank's sign segmentation on 400 brain MRI scans.
• The optimal model (U-net) was validated on two external datasets (300 scans each).
• Performance was assessed using Dice Similarity Coefficient (DSC) and Receiver Operating Characteristic (ROC) analysis.

Main Results:

• The U-net model achieved a DSC of 0.734 and an area under the ROC curve >0.9 on validation datasets.
• The model demonstrated high sensitivity, specificity, and accuracy in classifying Frank's sign.
• An intra-class correlation coefficient of 0.865 indicated strong reliability.

Conclusions:

• A robust deep learning model for automated Frank's sign segmentation from MRI scans was successfully developed.
• This automated tool shows significant potential to improve Frank's sign identification in clinical settings.
• The model can facilitate further research into the health implications of Frank's sign.