RESUMO
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.
RESUMO
OBJECTIVES: It has been reported that vascular plugging has become a therapeutic alternative to coil embolization in certain cases that require occlusion. However, the use of a relatively large and rigid delivery sheath may be a limiting factor in vascular plug use. In this study, we aimed to determine the safety and efficacy of a novel transcatheter occlusion device with unique design and delivery system in a pig model. METHODS: The Cera vascular plug was delivered and deployed through the FuStar steerable introducer sheath, which can control tip direction during advancement. Twelve pigs were randomized to undergo an embolization procedure in which the Cera vascular plug was implanted into the left internal iliac artery (IIA) with the FuStar steerable introducer (n = 6) or a control introducer sheath. Another eight pigs were assigned to undergo an embolization procedure in which the test device was implanted into either the splenic artery (SA, n = 4) or the lower segmental branch of left renal artery (LRA, n = 4). Angiography and pathological examinations were performed to evaluate the outcomes. RESULTS: A total of 20 target vessels were embolized with a total of 22 test plugs. Compared with the control introducer, plug embolization through the FuStar steerable introducer was associated with shorter fluoroscopy time (21.50 ± 3.62 vs. 28.33 ± 2.16 min, P = 0.003) and less contrast medium (129.17 ± 22.68 vs. 162.50 ± 13.69 mL, P = 0.012). At the 2-month follow-up, angiography and pathological examinations did not show any evidence of migration, and persistent occlusion was observed in 18 of the 20 target vessels. Organ ischemia occurred when plugs were deployed within the lower segmental branch of the LRA. CONCLUSION: This novel device is suitable for therapeutic vascular embolization with the use of flexible delivery systems. The different outcomes of SA and LRA plugging suggested that the occluding device should be placed within the appropriate portion of the target vessel to allow the development of collateralization.