Human identification plays a significant role in the investigations of disasters and criminal cases. Human identification could be achieved quickly and efficiently via 3D sphenoid sinus models by customized convolutional neural networks. In this retrospective study, a deep learning neural network was proposed to achieve human identification of 1475 noncontrast thin-slice CT scans. A total of 732 patients were retrieved and studied (82% for model training and 18% for testing). By establishing an individual recognition framework, the anonymous sphenoid sinus model was matched and cross-tested, and the performance of the framework also was evaluated on the test set using the recognition rate, ROC curve and identification speed. Finally, manual matching was performed based on the framework results in the test set. Out of a total of 732 subjects (mean age 46.45 years ± 14.92 (SD); 349 women), 600 subjects were trained, and 132 subjects were tested. The present automatic human identification has achieved Rank 1 and Rank 5 accuracy values of 93.94% and 99.24%, respectively, in the test set. In addition, all the identifications were completed within 55 s, which manifested the inference speed of the test set. We used the comparison results of the MVSS-Net to exclude sphenoid sinus models with low similarity and carried out traditional visual comparisons of the CT anatomical aspects of the sphenoid sinus of 132 individuals with an accuracy of 100%. The customized deep learning framework achieves reliable and fast human identification based on a 3D sphenoid sinus and can assist forensic radiologists in human identification accuracy.
International Journal of Legal Medicine - Short tandem repeats (STRs) are the preferred genetic markers in forensic DNA analysis, routinely measured by capillary electrophoresis (CE) method based... 相似文献
OBJECTIVE: At present, there are few materials available for esophagus reconstruction anywhere in the world. The reported survival rate in animals during the perioperative period is comparatively low. The present study assessed the feasibility of using a biotype artificial esophagus in the reconstruction of a dog's esophagus. METHODS: In 30 mongrel dogs, a portion of the thoracic esophagus was resected and an 8 cm section of artificial esophagus was transplanted to reconstruct the organ. The survival rate, food intake and process of healing were observed. RESULTS: Of the 30 dogs, 28 survived the perioperative period (93.3% survival). Two dogs (6.7%) developed an anastomotic fistula; 19 dogs survived for 1 year, a survival rate of 79.2% (19/24) with the remaining six dogs were killed according to the experimental protocol. Detachment of the artificial esophagus occurred on average 28.8 days after operation and the dogs suffered from varying degrees of dysphagia 23?45 days after operation. Gradual remission occurred after 4 months. The histological study revealed that the regenerated esophagus was composed of fibrous and connective tissues and the luminal surface was covered with squamous epithelium in 3?6 months. CONCLUSION: The transplanted artificial esophagus detached after the surrounding ‘regenerated esophagus’ had formed, and the squamous epithelium gradually covered the luminal surface. Continuous remodeling of the ‘regenerated esophagus’ gradually relieved the stenosis. Whether detachment of the implant and the postoperative stenosis can be solved is the key problem restricting the use of the biotype artificial esophagus in clinical practice. 相似文献
To define the gene content of the HLA class I region, cDNA selection was applied to three overlapping yeast artificial chromosomes (YACs) that spanned 1 megabase (Mb) of this region of the human major histocompatibility complex. These YACs extended from the region centromeric to HLA-E to the region telomeric to HLA-F. In addition to the recognized class I genes and pseudogenes and the anonymous non-class-I genes described recently by us and others, 20 additional anonymous cDNA clones were identified from this 1-Mb region. We also identified a long repetitive DNA element in the region between HLA-B and HLA-E. Homologues of this element were located at several sites in the human genome outside of the HLA complex. The portion of the HLA class I region represented by these YACs shows an average gene density as high as the class II and class III regions. Thus, the high gene density portion of the HLA complex is extended to more than 3 Mb. 相似文献