Ultrasound image analysis using deep learning algorithm for the diagnosis of thyroid nodules

Area of Science:

• Medical imaging
• Artificial intelligence in diagnostics
• Thyroid nodule evaluation

Background:

• Fine needle aspiration (FNA) is standard for thyroid nodule evaluation, but carries risks and costs.
• Nodules >2cm often require FNA, regardless of malignancy suspicion.
• There is a need for non-invasive methods to pre-emptively identify benign nodules.

Purpose of the Study:

• To develop and evaluate a deep learning image analysis model.
• To assess the model's ability to predict benign fine needle aspiration (FNA) results for thyroid nodules.
• To explore AI's potential in reducing unnecessary FNA procedures.

Main Methods:

• Retrospective collection of ultrasonographic thyroid nodule images with cytologic/histologic results.
• Training a deep learning model (Inception-V3) on 1358 nodule images (670 benign, 688 malignant).
• Validation using internal (n=55) and external (n=100) prospective test sets.

Main Results:

• Internal test set: 95.2% sensitivity for malignant nodules, 95.5% negative predictive value (NPV) for benign nodules.
• External test set: 94.0% sensitivity for malignant nodules, 90.3% NPV for benign nodules.
• The algorithm demonstrated high accuracy in classifying both benign and malignant thyroid nodules.

Conclusions:

• The deep learning algorithm shows promising sensitivity and NPV for thyroid nodule classification.
• Artificial intelligence can potentially assist clinicians in identifying nodules unlikely to be malignant.
• This AI approach may help avoid unnecessary fine needle aspiration (FNA) procedures.