The skin is the largest organ of the human body, and many visceral diseases will be directly reflected on the skin, so it is of great clinical significance to accurately segment the skin lesion images. To address the characteristics of complex color, blurred boundaries, and uneven scale information, a skin lesion image segmentation method based on dense atrous spatial pyramid pooling (DenseASPP) and attention mechanism is proposed. The method is based on the U-shaped network (U-Net). Firstly, a new encoder is redesigned to replace the ordinary convolutional stacking with a large number of residual connections, which can effectively retain key features even after expanding the network depth. Secondly, channel attention is fused with spatial attention, and residual connections are added so that the network can adaptively learn channel and spatial features of images. Finally, the DenseASPP module is introduced and redesigned to expand the perceptual field size and obtain multi-scale feature information. The algorithm proposed in this paper has obtained satisfactory results in the official public dataset of the International Skin Imaging Collaboration (ISIC 2016). The mean Intersection over Union (mIOU), sensitivity (SE), precision (PC), accuracy (ACC), and Dice coefficient (Dice) are 0.901 8, 0.945 9, 0.948 7, 0.968 1, 0.947 3, respectively. The experimental results demonstrate that the method in this paper can improve the segmentation effect of skin lesion images, and is expected to provide an auxiliary diagnosis for professional dermatologists.
ObjectiveTo discuss the security and effectiveness of fixing the unstable region of sagittal injured unit symmetrically with pedicle screws combined with bone graft fusion for treating thoracolumbar fractures.MethodsA series of 65 patients with a single level thoracolumbar fracture between November 2011 and November 2015 were included in the study. There were 41 males and 24 females with an average age of 36.7 years (range, 23-60 years). The fracture segments included T7 1 case, T9 in 2 cases, T10 in 4 cases, T11 in 8 cases, T12 in 14 cases, L1 in 19 cases, L2 in 13 cases, L3 in 3 cases, and L4 in 1 case. According to AO classification, there were 34 cases classified as type A, 27 cases type B, and 4 cases type C. The neurological function was evaluated by American Spinal Injury Association (ASIA) grade score, there were 1 case at grade A, 2 cases grade B, 6 cases grade C, 15 cases grade D, and 41 cases grade E. The thoracolumbar injury severity score (TLICS) was 4 in 9 cases, 5 in 29 cases, 6-8 in 23 cases, 9-10 in 4 cases. The time form injury to operation was 2-12 days (mean, 5.3 days). The fractured vertebra, along with the superior and inferior discs were defined as a injured unit and divided into three parts on the sagittal position: region Ⅰ mainly including the superior disc, cephalic 1/3 of injured vertebra, and posterior ligamentous complex as to oppose; region Ⅱ mainly including the middle 1/3 of injured vertebra, pedicles, lamina, spinous process, and supraspinal ligament; region Ⅲ mainly including the inferior disc, caudal 1/3 of injured vertebra, and posterior ligamentous complex as to oppose. The unstable region was defined as the key injured region of the vertebra. Pedicle screws were fixed symmetrically and correspondingly with bone grafting to treat thoracolumbar fractures. The neurological status, ratio of anterior body height, and sagittal Cobb angle were collected at preoperation, immediate after operation, and last follow-up to evaluate surgical and clinical outcomes.ResultsAll patients accepted operation safely and were followed up 12-24 months (mean, 17.3 months). Cerebrospinal fluid leakage occurred in 3 patients, and cured by symptomatic treatment. There was no complications such as loosening, displacement, and breakage of internal fixator. Bony fusion was achieved in all patients at 10-13 months (mean, 11.4 months) after operation. At last follow-up, according to ASIA grading, 1 case was grade A, 1 grade B, 3 grade C, 9 grade D, and 51 grade E, showing significant difference when compared with preoperative data (Z=–2.963, P=0.014). The ratio of anterior body height at preoperation, immediate after operation, and last follow-up were 53.2%±6.8%, 91.3%±8.3%, 89.5%±6.6% respectively; and the sagittal Cobb angle were (16.3±8.1), (2.6±7.5), (3.2±6.8)° respectively. The ratio of anterior body height and the sagittal Cobb angle at immediate after operation and at last follow-up were significantly improved when compared with preoperative values (P<0.05), but no significant difference was found between at immediate after operation and at last follow-up (P>0.05).ConclusionIt is safe and reliable to treat thoracolumbar fractures under the principle of fixing the unstable region of injured unit symmetrically with pedicle screws combined with bone grafting.
Objective To investigate the changes in spinal-pelvic sagittal parameters from preoperative standing to prone position in old traumatic spinal fractures with kyphosis. Methods The clinical data of 36 patients admitted between December 2016 and June 2021 for surgical treatment of old traumatic spinal fractures with kyphosis, including 7 males and 29 females, aged from 50 to 79 years (mean, 63.9 years), were retrospectively analyzed. Lesion segments included 2 cases of T11, 12 cases of T12, 2 cases of T11, 12, 4 cases of T12 and L1, 12 cases of L1, 2 cases of L2, 1 case of L2, 3, and 1 case of L3. The disease duration ranged from 4 to 120 months, with an average of 19.6 months. Surgical procedures included Smith-Petersen osteotomy in 4 cases, Ponte osteotomy in 6 cases, pedicle subtraction osteotomy in 2 cases, and improved fourth level osteotomy in 18 cases; the remaining 6 cases were not osteotomized. The bone mineral density ranged from −3.0 to 0.5 T, with a mean of −1.62 T. The spinal-pelvic sagittal parameters from preoperative standing to prone positions were measured, including local kyphosis Cobb angle (LKCA), thoracic kyphosis (TK), lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), and PI and LL mismatch (PI-LL). The kyphotic flexibility=(preoperative standing LKCA−preoperative prone LKCA)/preoperative standing LKCA×100%. Spinal-pelvic sagittal parameters were compared between standing position and prone position before operation, and Pearson correlation was used to judge the correlation between the parameters of standing position and prone position before operation. ResultsWhen the position changed from standing to prone, LKCA and TK decreased significantly (P<0.05), while SS, LL, PT, and PI-LL had no significant difference (P>0.05). Pearson correlation analysis showed that LL was significantly correlated with SS and PI-LL in both standing and prone positions (P<0.05), and the correlation strength between LL and SS in prone position was higher than that in standing position. In the standing position, LKCA was significantly correlated with SS and PT (P<0.05). However, when the position changed from standing to prone, the correlation between LKCA and SS and PT disappeared, while PT and PI-LL was positive correlation (P<0.05). The kyphotic flexibility was 25.13%-78.79%, with an average of 33.85%. Conclusion For the patients of old traumatic spinal fractures with kyphosis, the preoperative LKCA and TK decrease significantly from standing position to prone position, and the correlation between spinal and pelvic parameters also changed, which should be taken into account in the formulation of preoperative surgical plan.