This system is still very convenient to use. As a doctor, you don't need any extra steps. You just need to take a photo like a routine check-up and tell where the nodule is, what the outline is, and how benign and malignant it is. Generally speaking, it is still good, of course, it is also in the process of continuous updating and optimization.
The artificial intelligence mentioned by the doctor is a set of auxiliary diagnosis system for benign and malignant thyroid nodules based on ultrasonic images developed by Zhejiang Suntech Yunxing Image Technology Co., Ltd.
For the diagnosis of thyroid nodules, the average accuracy rate of doctors in top three hospitals is 60%-70%, and the primary hospitals will be lower. At present, the accuracy of the auxiliary diagnosis system can reach more than 85%. Ms. Hu Hairong, general manager of Suntech Yunxing, said that artificial intelligence is not qualified to practice medicine in China, so we positioned it as an auxiliary diagnosis and treatment, just to provide some reference for doctors. The future market positioning of this system is primary hospitals. After popularization, it can help grass-roots hospitals improve their diagnostic level, which is beneficial to the graded diagnosis and treatment policy implemented by the state.
It is understood that thyroid nodule refers to a lump in the thyroid gland, which can move up and down with the swallowing action. It is a common clinical disease, which can be caused by many reasons.
Most thyroid nodules are benign, but a few thyroid nodules are malignant from the beginning, and some nodules will change from benign to malignant. Therefore, once thyroid nodules are found in physical examination, doctors will suggest further examination, with the main purpose of distinguishing whether the nodules are "benign" or "malignant".
The proportion of thyroid nodules turning into cancer is 5%. If we can find them early and treat them as soon as possible, most patients can survive for a long time.
At present, the examination of thyroid nodules mainly depends on ultrasound. Compared with CT and MRI, ultrasound is cheap, flexible, real-time and has no side effects, and it is a popular detection method for ordinary people. However, its disadvantages are unclear images, large angiography and higher requirements for doctors. The advantages of CT and NMR are clear images, but the disadvantages are high detection cost and large side effects.
From the point of view that artificial intelligence is applied to medical image recognition or auxiliary diagnosis, ultrasound is more difficult than CT and magnetic resonance. We want to test the performance, efficiency and accuracy of the algorithm, so we choose ultrasound. At the same time, as a superficial organ, thyroid ultrasound is less complicated than abdomen and heart, and it is easier to make landing results.