This study aims to utilize advanced artificial intelligence (AI) image recog-nition technologies to establish a robust system for identifying features in lung computed tomog-raphy (CT) scans, thereby detecting respiratory infections such as SARS-CoV-2 pneumonia. Spe-cifically, the research focuses on developing a new model called Residual-Dense-Attention Gates U-Net (RDAG U-Net) to improve accuracy and efficiency in identification. : This study employed Attention U-Net, Attention Res U-Net, and the newly developed RDAG U-Net model.
View Article and Find Full Text PDFWe describe a rare case of a giant appendiceal mucocele as the leading point with ileocolic intussusception and the cause of acute right lower quadrant abdominal pain in an adult patient, which is correctly diagnosed by computed tomography (CT) imaging before surgery.
View Article and Find Full Text PDFBackground/purpose: Histologic diagnosis of suspicious small subpleural lung lesions is difficult and often impossible using existing image-guided needle biopsy techniques including video-assisted thoracoscopic wedge resection. Preoperative lung lesion localization provides a more obvious target to facilitate intraoperative resection. This study reviewed the indications, results and complications of CT-guided hook wire localization for subpleural lung lesions in video-assisted thoracoscopic surgery (VATS).
View Article and Find Full Text PDFThe Escherichia coli AlkB protein was recently found to repair cytotoxic DNA lesions 1-methyladenine and 3-methylcytosine by using a novel iron-catalyzed oxidative demethylation mechanism. Three human homologs, ABH1, ABH2 and ABH3, have been identified, and two of them, ABH2 and ABH3, were shown to have similar repair activities to E.coli AlkB.
View Article and Find Full Text PDF