Sino-Singapore Online, June 22 (Chen Jing, Wang Genhua) The detection rate of lung nodules increased from 77.3% to 89.3%, the accuracy of early lung cancer diagnosis reached 94.7%, and the accuracy of invasive judgment of pure ground glass density nodules was 93% ...
The reporter was informed on the 22nd that the radiodiagnostic team led by Professor Liu Shiyuan, the Second Affiliated Hospital of Naval Military Medical University (Shanghai Changzheng Hospital), built an early screening and early diagnosis imaging system for lung cancer, formed an early lung cancer diagnosis standard and standard, and proposed a series of rich pioneering Sex theory and original diagnostic methods, thereby improving the early diagnosis rate of lung cancer and reducing mortality.
It is reported that the five-year survival rate of early lung cancer after surgical resection is close to 100%. According to Liu Shiyuan, unfortunately, many patients have symptoms of discomfort until they find out that they have lung cancer. At this time, about 80% of patients have lost the opportunity for surgery, and less than 2% of lung cancers are in the early stage when they are discovered.
Liu Shiyuan told reporters that for more than ten years, the research team has carried out prospective exploration and research on the worldwide problems of difficult diagnosis, high misdiagnosis rate and low diagnosis rate of lung cancer diagnosis and treatment. It is reported that the relevant research won the first prize of 2019 Shanghai Science and Technology Progress Award.
The early diagnosis is closely related to the prognosis of lung cancer. The research team took the lead in carrying out large-scale multi-center early lung cancer screening in Shanghai, and conducted low-dose CT examination of the high-risk populations initially screened. Expandable lung cancer case database to achieve sharing of case data.
"Some people do lung CT scans, and when they get the report, they only look at lung nodules. They often feel helpless and think they have lung cancer." Liu Shiyuan said that lung nodules are not necessarily lung cancer. How can we tell if the nodule found is not lung cancer? Radiologists can dig out more diagnostic information from the images.
The research team and the artificial intelligence company jointly developed artificial intelligence models. The earliest internationally used artificial lung nodule detection model was used for chest radiography and CT for related examinations. Not only can detect the nodule size of the millimeter of the needle tip, but also can be marked The location and density of the nodules avoid the risk of missed diagnosis.
Whether or not pure ground glass nodules require surgery has always been a problem faced by surgery. The research team built a variety of accurate prediction models and used a variety of methods to judge. According to statistics, the accuracy of this method is over 93%. The research group also innovatively used targeted probes to inject a contrast agent imaging system, like the "miracle mirror", to quickly "present" the malignant tumor and help the doctor to "precisely extract bad molecules".
It is understood that many original achievements of Liu Shiyuan's team, such as pulmonary nodule examination technology, diagnosis and differential diagnosis, and lung cancer screening, have been widely used in Shanghai and hundreds of major hospitals nationwide. Pulmonary nodule report analysis system, benefiting patients all over the world.
In recent years, Liu Shiyuan led the team to formulate 5 Chinese expert consensuses, formed 2 industry norms and standards, and obtained 7 national invention patents and 3 utility model patents.