Application Of Dual Artificial Intelligence In The Diagnosis And Treatment Of Multiple Pulmonary Nodules

Objective: To evaluate the application value of artificial intelligence image-assisted diagnosis system and three-dimensional computed tomography bronchography and angiography system in the diagnosis and treatment of multiple pulmonary nodules.Methods: A total of 134 patients with multiple lung nodules who underwent television-assisted thoracoscopic surgery in the first ward of the Department of Thoracic Surgery of the First People’s Hospital of Jining,Shandong Province from August 2019 to August 2021.Firstly,the preoperative chest CT data of patients with multiple lung nodules were screened and sorted out by AI-assisted diagnosis system and manual reading screening,and the nodules were matched with paraffin pathology results to analyze the ability of AI-assisted diagnostic system in the screening of multiple pulmonary nodules.Secondly,the patients were divided into the experimental group and the control group according to whether the patients underwent preoperative 3D-CTBA.The preoperative clinical data of the two groups of patients were analyzed by propensity score matching,and the perioperative indicators of the two groups of patients after matching were analyzed.At the same time,the perioperative indicators of the patients in the experimental group and the control group who underwent segmentectomy were retrospectively analyzed to further evaluate the value of3D-CTBA in the treatment of multiple pulmonary nodules.Finally,the treatment of high-risk nodules after MPNs surgery was explored.Results: The enrolled patients were more common in women（67.91%）and people without smoking history（79.10%）.80 patients were diagnosed with multiple primary lung cancers,and the pathological combination was mainly adenocarcinoma-adenocarcinoma（96.25%）.In terms of screening,compared with the manual reading group,the AI-assisted diagnostic system was more sensitive in screening for micronodular with a diameter of ≤5 mm（P<0.05）,and the sensitivity of screening for multiple primary lung cancers was better than the manual reading group（87.50%vs.50.00%,P=0.043）.In terms of preoperative planning,3D-CTBA is able to identify all abnormal anatomical vasculatures preoperatively.After matching patients with sex,age,FEV1%,and BMI propensity,the 3DCTBA group had less postoperative drainage（P=0.016）.Among the perioperative indicators of the two groups of patients who underwent segmentectomy,the 3D-CTBA group had advantages in terms of total drainage,duration of surgery,and control of postoperative complications（P<0.05）.Conclusion: Artificial intelligence software has certain auxiliary value in the diagnosis and treatment of multiple lung nodules,but it is still necessary to increase the sample size for further verification.In addition,patients with different multiple lung nodules still need to be individualized treatment plans by surgeons.