Résumé
Precise segmentation of lung field is a crucial step in chest radiographic computer-aided diagnosis system. With the development of deep learning, fully convolutional network based models for lung field segmentation have achieved great effect but are poor at accurate identification of the boundary and preserving lung field consistency. To solve this problem, this paper proposed a lung segmentation algorithm based on non-local attention and multi-task learning. Firstly, an encoder-decoder convolutional network based on residual connection was used to extract multi-scale context and predict the boundary of lung. Secondly, a non-local attention mechanism to capture the long-range dependencies between pixels in the boundary regions and global context was proposed to enrich feature of inconsistent region. Thirdly, a multi-task learning to predict lung field based on the enriched feature was conducted. Finally, experiments to evaluate this algorithm were performed on JSRT and Montgomery dataset. The maximum improvement of Dice coefficient and accuracy were 1.99% and 2.27%, respectively, comparing with other representative algorithms. Results show that by enhancing the attention of boundary, this algorithm can improve the accuracy and reduce false segmentation.
Sujets)
Rayons X , Algorithmes , Diagnostic assisté par ordinateur , Thorax/imagerie diagnostique , Poumon/imagerie diagnostique , Traitement d'image par ordinateurRésumé
In recent years, epileptic seizure detection based on electroencephalogram (EEG) has attracted the widespread attention of the academic. However, it is difficult to collect data from epileptic seizure, and it is easy to cause over fitting phenomenon under the condition of few training data. In order to solve this problem, this paper took the CHB-MIT epilepsy EEG dataset from Boston Children's Hospital as the research object, and applied wavelet transform for data augmentation by setting different wavelet transform scale factors. In addition, by combining deep learning, ensemble learning, transfer learning and other methods, an epilepsy detection method with high accuracy for specific epilepsy patients was proposed under the condition of insufficient learning samples. In test, the wavelet transform scale factors 2, 4 and 8 were set for experimental comparison and verification. When the wavelet scale factor was 8, the average accuracy, average sensitivity and average specificity was 95.47%, 93.89% and 96.48%, respectively. Through comparative experiments with recent relevant literatures, the advantages of the proposed method were verified. Our results might provide reference for the clinical application of epilepsy detection.
Sujets)
Enfant , Humains , Algorithmes , Apprentissage profond , Électroencéphalographie , Épilepsie/diagnostic , Crises épileptiques/diagnostic , Traitement du signal assisté par ordinateur , Analyse en ondelettesRésumé
Speech feature learning is the core and key of speech recognition method for mental illness. Deep feature learning can automatically extract speech features, but it is limited by the problem of small samples. Traditional feature extraction (original features) can avoid the impact of small samples, but it relies heavily on experience and is poorly adaptive. To solve this problem, this paper proposes a deep embedded hybrid feature sparse stack autoencoder manifold ensemble algorithm. Firstly, based on the prior knowledge, the psychotic speech features are extracted, and the original features are constructed. Secondly, the original features are embedded in the sparse stack autoencoder (deep network), and the output of the hidden layer is filtered to enhance the complementarity between the deep features and the original features. Third, the L1 regularization feature selection mechanism is designed to compress the dimensions of the mixed feature set composed of deep features and original features. Finally, a weighted local preserving projection algorithm and an ensemble learning mechanism are designed, and a manifold projection classifier ensemble model is constructed, which further improves the classification stability of feature fusion under small samples. In addition, this paper designs a medium-to-large-scale psychotic speech collection program for the first time, collects and constructs a large-scale Chinese psychotic speech database for the verification of psychotic speech recognition algorithms. The experimental results show that the main innovation of the algorithm is effective, and the classification accuracy is better than other representative algorithms, and the maximum improvement is 3.3%. In conclusion, this paper proposes a new method of psychotic speech recognition based on embedded mixed sparse stack autoencoder and manifold ensemble, which effectively improves the recognition rate of psychotic speech.
Sujets)
Humains , Algorithmes , Bases de données factuelles , Troubles psychotiques , Parole , Perception de la paroleRésumé
Objective To investigate the prevalence of genital Chlamydia trachomatis (GCT) infection and associated risk factors in male clients attending sexually transmitted disease (STD) clinics in Guangdong and provide integrated intervention strategy for this group.Methods Convenient sampling was used to recruit participants from April to June in 2015 in Guangdong province.The information about their socio-demographic characteristics and sexual behaviors were collected by using a questionnaire,and blood samples were taken from them to test the antibodies against HIV,syphilis and HCV.First pass urine was taken to test GCT and gonorrhea.Results A total of 1 749 participants with the average age of 39.53 years were recruited.The majority of them were married (73.87%,1 292/1 749),residents of Guangdong (92.28%,1 614/1 749) and in Han ethnic group (99.49%,1 740/1 749).The positive rates for GCT,HIV,syphilis,HCV,Neisseria gonorrhea,and WBC in urinalysis were 6.06% (106/1 749),0.46% (8/1 749),3.43% (60/1 749),0.45% (7/1 550),2.74% (48/1 749),7.89% (138/1 749) respectively.Multivariate analysis showed that risk factors for GCT infection include IDUs (OR=13.98,95%CI:3.35-58.38),anal sex with men (OR=3.11,95% CI:1.45-6.71),Neisseria gonorrhea positive (OR =9.64,95% CI:5.09-18.24),and WBC positive (OR =1.96,95% CI:1.08-3.55).Conclusions This study demonstrated the high prevalence of GCT infection in male clients attending STD clinics in Guangdong.Therefore precision intervention should target this population at high-risk.
Résumé
Objective To investigate the prevalence of genital Chlamydia trachomatis (GCT) infection and associated risk factors in male clients attending sexually transmitted disease (STD) clinics in Guangdong and provide integrated intervention strategy for this group.Methods Convenient sampling was used to recruit participants from April to June in 2015 in Guangdong province.The information about their socio-demographic characteristics and sexual behaviors were collected by using a questionnaire,and blood samples were taken from them to test the antibodies against HIV,syphilis and HCV.First pass urine was taken to test GCT and gonorrhea.Results A total of 1 749 participants with the average age of 39.53 years were recruited.The majority of them were married (73.87%,1 292/1 749),residents of Guangdong (92.28%,1 614/1 749) and in Han ethnic group (99.49%,1 740/1 749).The positive rates for GCT,HIV,syphilis,HCV,Neisseria gonorrhea,and WBC in urinalysis were 6.06% (106/1 749),0.46% (8/1 749),3.43% (60/1 749),0.45% (7/1 550),2.74% (48/1 749),7.89% (138/1 749) respectively.Multivariate analysis showed that risk factors for GCT infection include IDUs (OR=13.98,95%CI:3.35-58.38),anal sex with men (OR=3.11,95% CI:1.45-6.71),Neisseria gonorrhea positive (OR =9.64,95% CI:5.09-18.24),and WBC positive (OR =1.96,95% CI:1.08-3.55).Conclusions This study demonstrated the high prevalence of GCT infection in male clients attending STD clinics in Guangdong.Therefore precision intervention should target this population at high-risk.
Résumé
Objective To clone ,construction ,express and purify Tp47 of Treponema pallidum (Tp) ,and assess the immunoreac‐tivity by Western‐Blot .Methods Tp47 was amplified by polymerase chain reaction ,and then cloned to the vector pGEX‐6P‐1 .The correct sequence of the recombinant plasmids pGEX‐6P1‐Tp47 was transformed into Escherichia coli BL21 (DE3) and induced .The expression product was analyzed by sodium dodecyl sulfate polyacrylamide gel electropheresis and Western‐Blot .The expression protein was purified .Serum of different clinical stages of syphilis was used as the antibody to detect the immunoreactivity of the protein by Western‐Blot .Results A fusion protein with molecular weight about 71 × 103 was attained .Western‐Blot proved that the recombinant protein can react with Tp IgG positive sera .And the specificities and sensitivities of the diagnostic reagent detected by sera were 100% .Conclusion The recombinant protein Tp47 was expressed and purified with good antigen activity ,which could provide the basis of theory and practice for the development of early diagnostic kit applying to detect Tp infection .
Résumé
ObjectiveTo analysis the clinic outcome(similarities & differences, indication and caution)of repairing fingertip soft tissue defect with two kind of reverse digital artery island flaps with the palmar cutaneous branch of proper digital nerve.MethodsFrom January 2000 to Auguest 2008, all 500cases (514 fingers) were repaired with reverse homodigital artery island flap (263 fingers of 250 cases) and reverse cross finger digital artery island flap(251 fingers of 250 cases).All cases were followed up from 3-8years,compared with flap range,finger extension,sensory rehabilitation,et al. ResultsAll flaps (A & Bgroup) of 500 cases (514 fingers) survived.According to flap range,cross finger flap was better than homodigital flap (P < 0.05); according to finger extension,homodigital flap better than cross finger flap (P <0.05); according to sensory rehabilitation,homodigital flap little better than cross finger flap(P < 0.05); according to long-term follow-up, all had no obviously different (P > 0.05).ConclusionHomodigital flap can cover distal soft tissue defect less than 2.0 cm× 2.5 cm, cross finger flap is a better choice for more than 2.0cm×3.0cm of distal soft tissue defect. According to the flap cover area, cross finger flap is first choice,and to finger extension & sensory rehabilitation,homodigital flap is first choice; accoding to finger movements,two kinds of flaps are all good choice.