In the past decade, the incidence rate of thyroid cancer has increased worldwide and has become a focus of public safety. Two of the most common reasons for this are the popularization of simple cancer detection technologies and the enhancement of people’s awareness of physical examinations.
Among the many options available for thyroid cancer diagnosis and treatment, FNAB (Fine-needle aspiration biopsy) stands out as a standard choice for its reliable accuracy and excellent diagnostic performance. However, as they come with certain complications, thyroid specialists now say that this procedure can be ideally replaced with other underrated diagnosis options like AI-based thyroid imaging. They also believe that the need for FNA for further examinations can be decided upon the results of AI-based thyroid imaging tests or due to other rare conditions.
Here is more on how the AI model for thyroid imaging can avoid unnecessary biopsies and surgeries
Most Thyroid Nodules Are Noncancerous
While thyroid nodules are common, only a small percentage are actually cancerous. Statistically speaking, more than 95% — of thyroid nodules are benign (noncancerous). This alone rules out the need to conduct relatively invasive procedures like Fine Needle Aspiration Biopsy as a first line of defense and instead opt for other simpler options.
Although ultrasound is a preferred imaging examination to evaluate benign and malignant thyroid nodules, the diagnostic results are affected by factors such as the personal experience of the thyroid specialist or ultrasound doctor, equipment used, operating skills, basic thyroid lesions, and the diagnostic accuracy and efficiency of doctors at different levels and with different seniority vary greatly. These drawbacks contributed as a driving factor that made FNA more preferable and reliable.
As a result, these issues stimulated the development of AI solutions to assist physicians with ultrasound imaging and make accurate diagnoses and judgments on treatment options.
A new study presented at ENDO 2022, the Endocrine Society’s annual meeting, reports that artificial intelligence (AI) can be used effectively to identify thyroid nodules seen on thyroid ultrasound that are very unlikely to be cancerous, thereby significantly reducing a large number of unnecessary biopsies.
According to the study lead researcher Nikita Pozdeyev, MD, Ph.D., of the University of Colorado Anschutz Medical Campus in Aurora, the AI-based model achieves a sensitivity of 97% in cancer detection and a speciﬁcity of 61% to detect cancer correctly. This study also demonstrates that the ultrasound-based AI classiﬁer of thyroid nodules achieves sensitivity comparable to that of thyroid biopsy with ﬁne needle aspiration. The technology can assist radiologists and endocrinologists in choosing which thyroid nodules should undergo biopsy more precisely.
Another similar study presented at the 2022 Multidisciplinary Head and Neck Cancers Symposium concluded that integrating different AI methods made it possible to capture more data while minimizing noise, achieving a high level of accuracy in making predictions. It also revealed that a properly designed multimodal model comprising radiomics, TDA, and (ML)TI-RADS was able to distinguish pathological stages (93% accuracy for T-stage, 89% for N-stage, and 98% for extrathyroidal extension).
AI In Endocrinology Management
For those new to AI technology, in a broader sense, it represents a machine or computer that can replicate the competencies and expertise of a human brain and, therefore, can think, process, and make decisions based on past learned experiences. In 2021, AI in the healthcare market was worth over 11 billion U.S. dollars worldwide and is forecasted to reach around 188 billion U.S. dollars by 2030. Further, as the importance of this technology is realized in diagnosing malignant thyroid nodules, it is also expected to play a prominent role in the future of endocrinology management.
At GluCare, our thyroid image scanning uses a unique AI (Artificial Intelligence) diagnosis approach to evaluate and identify the sonographic calcification of thyroid nodules during the same visit.
The GluCare AI-based image-scanning model has been validated through numerous studies and has till now proven highly accurate in detecting cancerous or benign thyroid nodules.
We also have the expertise and technology to analyze thyroid images more accurately and efficiently than conventional methods. This makes it easy for us to make informed treatment decisions quickly while spending more time interacting with patients and building an excellent doctor-patient relationship (DPR).