A novel deep-learning based weighted feature fusion architecture for precise classification of pressure injury.

Introduction: Precise classification has an important role in treatment of pressure injury (PI), while current machine-learning or deeplearning based methods of PI classification remain low accuracy. Methods: In this study, we developed a deeplearning based weighted feature fusion architecture for fine-grained classification, which combines a top-down and bottom-up pathway to fuse high-level semantic information and low-level detail representation. We validated it in our established database that consist of 1,519 images from multi-center clinical cohorts. ResNeXt was set as the backbone network. Results: We increased the accuracy of stage 3 PI from 60.3% to 76.2% by adding weighted feature pyramid network (wFPN). The accuracy for stage 1, 2, 4 PI were 0.870, 0.788, and 0.845 respectively. We found the overall accuracy, precision, recall, and F1-score of our network were 0.815, 0.808, 0.816, and 0.811 respectively. The area under the receiver operating characteristic curve was 0.940. Conclusions: Compared with current reported study, our network significantly increased the overall accuracy from 75% to 81.5% and showed great performance in predicting each stage. Upon further validation, our study will pave the path to the clinical application of our network in PI management.

Effect of dexmedetomidine on somatosensory- and motor-evoked potentials in patients receiving craniotomy under propofol-sevoflurane combined anesthesia.

Introduction: Dexmedetomidine is often used as an adjunct to total intravenous anesthesia (TIVA) for procedures requiring intraoperative neurophysiologic monitoring (IONM). However, it has been reported that dexmedetomidine might mask the warning of a neurological deficit on intraoperative monitoring. Methods: We reviewed the intraoperative neurophysiological monitoring data of 47 patients who underwent surgery and IONM from March 2019 to March 2021 at the Department of Neurosurgery, Renmin Hospital of Wuhan University. Pre- and postoperative motor function scores were recorded and analyzed. Dexmedetomidine was administered intravenously at 0.5 µg/kg/h 40 min after anesthesia and discontinued after 1 h in the dexmedetomidine group. Results: We found that the amplitude of transcranial motor-evoked potentials (Tce-MEPs) was significantly lower in the dexmedetomidine group than in the negative control group (P < 0.0001). There was no statistically significant difference in the somatosensory-evoked potentials (SSEPs) amplitude or the Tce-MEPs or SSEPs latency. There was no significant decrease in postoperative motor function in the dexmedetomidine group compared with the preoperative group, suggesting that there is no evidence that dexmedetomidine affects patient prognosis. In addition, we noticed a synchronized bilateral decrease in the Tce-MEPs amplitude in the dexmedetomidine group and a mostly unilateral decrease on the side of the brain injury in the positive control group (P = 0.001). Discussion: Although dexmedetomidine does not affect the prognosis of patients undergoing craniotomy, the potential risks and benefits of applying it as an adjunctive medication during craniotomy should be carefully evaluated. When dexmedetomidine is administered, Tce-MEPs should be monitored. When a decrease in the Tce-MEPs amplitude is detected, the cause of the decrease in the MEPs amplitude can be indirectly determined by whether the decrease is bilateral.