AIBx, artificial intelligence model to risk stratify thyroid nodules

Abstract

Background Current classification systems for thyroid nodules are very subjective. Artificial intelligence (AI) algorithms have been used to decrease subjectivity in medical image interpretation. 1 out of 2 women over the age of 50 may have a thyroid nodule and at present the only way to exclude malignancy is through invasive procedures. Hence, there exists a need for noninvasive objective classification of thyroid nodules. Some cancers have benign appearance on ultrasonogram. Hence, we decided to create an image similarity algorithm rather than image classification algorithm. Methods Ultrasound images of thyroid nodules from patients who underwent either biopsy or thyroid surgery from February of 2012 through February of 2017 in our institution were used to create AI models. Nodules were excluded if there was no definitive diagnosis of benignity or malignancy. 482 nodules met the inclusion criteria and all available images from these nodules were used to create the AI models. Later, these AI models were used to test 103 thyroid nodules which underwent biopsy or surgery from March of 2017 through July of 2018. Results Negative predictive value of the image similarity model was 93.2%. Sensitivity, specificity, positive predictive value and accuracy of the model was 87.8%, 78.5%, 65.9% and 81.5% respectively. Conclusion When compared to published results of ACR TIRADS and ATA classification system, our image similarity model had comparable negative predictive value with better sensitivity specificity and positive predictive value. By using image similarity AI models, we can eliminate subjectivity and decrease the number of unnecessary biopsies. Using image similarity AI model, we were able to create an explainable AI model which increases physician′s confidence in the predictions.

Thyroid FNA terminology: The case for a single unified international system for thyroid FNA reporting

“…This will be particularly important in the context of clinical trials, evaluations of peer‐reviewed literature, or commercial studies. Examples include development or introduction of new molecular methods, validation of artificial intelligence routines for thyroid nodule assessment using ultrasound and histopathology, eg for assessment of ultrasound characteristics, 36 or morphometry of papillary carcinoma‐type nuclei to ascertain whether a particular lesion is benign or malignant 37 . It will therefore be essential that molecular pathology methods, histopathology terminologies, and cytopathology terminologies are very much aligned so that the clinical results obtained in one part of the world can be extrapolated with ease and utilised in other parts of the world.…”