In order to solve the problems that the current synthetic aperture radar(SAR)image target detection method cannot adapt to targets of different sizes,and the complex image background leads to low detection accuracy,an...In order to solve the problems that the current synthetic aperture radar(SAR)image target detection method cannot adapt to targets of different sizes,and the complex image background leads to low detection accuracy,an improved SAR image small target detection method based on YOLOv7 was proposed in this study.The proposed method improved the feature extraction network by using Switchable Around Convolution(SAConv)in the backbone network to help the model capture target information at different scales,thus improving the feature extraction ability for small targets.Based on the attention mechanism,the DyHead module was embedded in the target detection head to reduce the impact of complex background,and better focus on the small targets.In addition,the NWD loss function was introduced and combined with CIoU loss.Compared to the CIoU loss function typically used in YOLOv7,the NWD loss function pays more attention to the processing of small targets,so as to further improve the detection ability of small targets.The experimental results on the HRSID dataset indicate that the proposed method achieved mAP@0.5 and mAP@0.95 scores of 93.5%and 71.5%,respectively.Compared to the baseline model,this represents an increase of 7.2%and 7.6%,respectively.The proposed method can effectively complete the task of SAR image small target detection.展开更多
AIM： TO evaluate quality of life （QOL） following Ivor Lewis, left transthoracic, and combined thoracoscopic/ laparoscopic esophagectomy in patients with esopha- geal cancer. METHODS： Ninety patients with esophagea...AIM： TO evaluate quality of life （QOL） following Ivor Lewis, left transthoracic, and combined thoracoscopic/ laparoscopic esophagectomy in patients with esopha- geal cancer. METHODS： Ninety patients with esophageal cancer were assigned to Ivor Lewis （/7 = 30）, combined thora- coscopic/laparoscopic （n = 30）, and left transthoracic （n = 30） esophagectomy groups. The QOL-core 30 questionnaire and the supplemental QOL-esophageal module 18 questionnaire for patients with esophageal cancer, both developed by the European Organization for Research and Treatment of Cancer, were used to evaluate patients＇ QOL from 1 wk before to 24 wk after surgery. RESULTS： A total of 324 questionnaires were collect- ed from 90 patients, 36 postoperative questionnaires were not completed because patients could not be contacted for follow-up visits. QOL declined markedly in all patients at 1 wk postoperatively： preoperative and 1-wk postoperative global QOL scores in the Ivor Lewis, combined thoracoscopic/laparoscopic, and left transthoracic groups were 80.8 ± 9.3 vs 32.0 ± 16.1 （P 〈 0.001）, 81.1±9.0 vs 53.3 ± 11.5 （P 〈 0.001）, and 83.6 ± 11.2 vs 46.4 ± 11.3 （P 〈 0.001）, respectively. Thereafter, QOL recovered gradually in all patients. Patients who underwent Ivor Lewis esophagectomy showed the most pronounced decline in QOL; global scores were lower in this group than in the combined thoracoscopic/laparoscopic （P 〈 0.001） and left trans- thoracic （P 〈 0.001） groups at 1 wk postoperatively and was not restored to the preoperative level at 24 wk postoperatively. QOL declined least in patients under- going combined thoracoscopic/laparoscopic esopha- gectomy, and most indices had recovered to preopera- tive levels at 24 wk postoperatively. In the Ivor Lewis and combined thoracoscopic/laparoscopic groups, pain and physical function scores were 78.9 ± 18.5 vs 57.8 ± 19.9 （P 〈 0.001） and 59.3 ± 16.1 vs 70.2 ± 19.2 （P = 0.02）, respectively, at 1 wk postoperatively and 26.1 ± 28.6 vs 9.5 ± 15.6 （P = 0.007） and 88.4 ± 10.5 vs 95.8 ± 7.3 （P = 0.003）, respectively, at 24 wk postop- eratively. Scores in the left transthoracic esophagecto- my group fell between those of the other two groups. CONCLUSION： Compared with Ivor Lewis and left transthoracic esophagectomies, combined thoraco- scopic/laparoscopic esophagectomy enables higher postoperative QOL, making it a preferable surgical ap- proach for esophageal cancer.展开更多
文摘In order to solve the problems that the current synthetic aperture radar(SAR)image target detection method cannot adapt to targets of different sizes,and the complex image background leads to low detection accuracy,an improved SAR image small target detection method based on YOLOv7 was proposed in this study.The proposed method improved the feature extraction network by using Switchable Around Convolution(SAConv)in the backbone network to help the model capture target information at different scales,thus improving the feature extraction ability for small targets.Based on the attention mechanism,the DyHead module was embedded in the target detection head to reduce the impact of complex background,and better focus on the small targets.In addition,the NWD loss function was introduced and combined with CIoU loss.Compared to the CIoU loss function typically used in YOLOv7,the NWD loss function pays more attention to the processing of small targets,so as to further improve the detection ability of small targets.The experimental results on the HRSID dataset indicate that the proposed method achieved mAP@0.5 and mAP@0.95 scores of 93.5%and 71.5%,respectively.Compared to the baseline model,this represents an increase of 7.2%and 7.6%,respectively.The proposed method can effectively complete the task of SAR image small target detection.
文摘AIM： TO evaluate quality of life （QOL） following Ivor Lewis, left transthoracic, and combined thoracoscopic/ laparoscopic esophagectomy in patients with esopha- geal cancer. METHODS： Ninety patients with esophageal cancer were assigned to Ivor Lewis （/7 = 30）, combined thora- coscopic/laparoscopic （n = 30）, and left transthoracic （n = 30） esophagectomy groups. The QOL-core 30 questionnaire and the supplemental QOL-esophageal module 18 questionnaire for patients with esophageal cancer, both developed by the European Organization for Research and Treatment of Cancer, were used to evaluate patients＇ QOL from 1 wk before to 24 wk after surgery. RESULTS： A total of 324 questionnaires were collect- ed from 90 patients, 36 postoperative questionnaires were not completed because patients could not be contacted for follow-up visits. QOL declined markedly in all patients at 1 wk postoperatively： preoperative and 1-wk postoperative global QOL scores in the Ivor Lewis, combined thoracoscopic/laparoscopic, and left transthoracic groups were 80.8 ± 9.3 vs 32.0 ± 16.1 （P 〈 0.001）, 81.1±9.0 vs 53.3 ± 11.5 （P 〈 0.001）, and 83.6 ± 11.2 vs 46.4 ± 11.3 （P 〈 0.001）, respectively. Thereafter, QOL recovered gradually in all patients. Patients who underwent Ivor Lewis esophagectomy showed the most pronounced decline in QOL; global scores were lower in this group than in the combined thoracoscopic/laparoscopic （P 〈 0.001） and left trans- thoracic （P 〈 0.001） groups at 1 wk postoperatively and was not restored to the preoperative level at 24 wk postoperatively. QOL declined least in patients under- going combined thoracoscopic/laparoscopic esopha- gectomy, and most indices had recovered to preopera- tive levels at 24 wk postoperatively. In the Ivor Lewis and combined thoracoscopic/laparoscopic groups, pain and physical function scores were 78.9 ± 18.5 vs 57.8 ± 19.9 （P 〈 0.001） and 59.3 ± 16.1 vs 70.2 ± 19.2 （P = 0.02）, respectively, at 1 wk postoperatively and 26.1 ± 28.6 vs 9.5 ± 15.6 （P = 0.007） and 88.4 ± 10.5 vs 95.8 ± 7.3 （P = 0.003）, respectively, at 24 wk postop- eratively. Scores in the left transthoracic esophagecto- my group fell between those of the other two groups. CONCLUSION： Compared with Ivor Lewis and left transthoracic esophagectomies, combined thoraco- scopic/laparoscopic esophagectomy enables higher postoperative QOL, making it a preferable surgical ap- proach for esophageal cancer.
