A multimodal artificial intelligence (AI) platform can accurately identify thyroid cancer using routine ultrasound images, according to research presented at the 2022 Multidisciplinary Head and Neck Cancers Symposium.1
“We have developed an artificial intelligence platform that would examine ultrasound images and predict with high accuracy whether a potentially problematic thyroid nodule is, in fact, cancerous,” study author Annie Chan, MD, of Mass General Cancer Center in Boston, said in a statement.2
“If it is cancerous, we can further predict the tumor stage, the nodal stage, and the presence or absence of BRAF mutation,” Dr Chan added.
The platform combines multiple AI methods, including radiomics, topological data analysis, deep learning, and machine learning.
“By integrating different AI methods, we were able to capture more data while minimizing noise,” Dr Chan explained. “This allows us to achieve a high level of accuracy in making predictions.”
The AI platform was tested on 1346 ultrasound images from 784 patients with benign or malignant thyroid nodules. The images were divided into an internal training and validation dataset (156 malignant and 357 benign nodules) and an external validation dataset (270 malignant and 50 benign nodules).
The AI platform predicted thyroid cancer with an accuracy of 98% in the internal dataset and 93% in the external dataset.
Dr Chan noted that the multimodal AI platform was superior to deep learning alone with regard to accuracy (98% vs 87%), precision (96% vs 90%), sensitivity (98% vs 72%), and specificity (98% vs 96%).
The AI platform was able to distinguish pathological stage as well, with a 93% accuracy for T stage, 90% accuracy for N stage, and 98% accuracy for extracapsular extension. The platform was able to detect the BRAF V600E mutation with an accuracy of 96%.
1. Chan AW, Paul R, Juliano A, Faquin W. An artificial intelligence ultrasound platform for screening and staging of thyroid cancer. Multidisciplinary Head and Neck Cancers Symposium 2022. February 24-26, 2022. Abstract 10.
2. Artificial intelligence platform shows potential for thyroid cancer screening and staging from ultrasound images. ASTRO. News release. Published February 24, 2022. Accessed February 25, 2022. https://www.astro.org/News-and-Publications/News-and-Media-Center/News-Releases/2022/Artificial-intelligence-platform-shows-potential-f