Ultrasound imaging guided with artificial intelligence (AI) noninvasively detects almost all malignant thyroid nodules, researchers reported at the 2022 Multidisciplinary Head and Neck Cancers Symposium. It is also accurate when predicting T stage, extracapsular extension, and presence of a BRAF variant.

The researchers used the AI system to evaluate 1,346 ultrasound images of 784 patients with known thyroid nodule status, including internal training and validation studies (156 malignant and 357 benign lesions) and an external validation study (270 malignant and 50 benign nodules).

The AI system has four different components: thyroid imaging reporting and data system (TI-RADS), radiomics (quantifies texture and differing gray level intensity in an image), topologic data analysis (TDA), and deep learning. The researchers found that radiomics was the most accurate at predicting malignancy (88.7%), followed by deep learning (87.4%), TDA (81.5%), and TI-RADS (80%). The system was also 93% accurate for predicting T stage, 98% accurate for extracapsular extension, and 96% accurate for BRAF variants.

“We need to minimize subjectivity,” the researchers said. “We need to be able to make a more precise and accurate, as well as quantitative, prediction. We need to distinguish real versus fake. We need to remove noises from the images. We need to remove noises from our data.”