A converged network based on deep learning is proposed to realize the efficient and accurate diagnosis of diabetic retinopathy.Both data augmentation technology and generative adversarial network are used to augment the fundus images in EyePACS dataset for effectively addressing the problem of uneven classification of fundus images.The proposed model uses Inception-Resnet-V2 as the main network,and incorporates deep residual shrinkage network and pyramid split attention module for effectively filtering out the irrelevant information in the feature learning process and focusing on the lesion information,thereby improving the network's ability to capture important features.Experimental results show that the optimized model achieves accuracy,recall,specificity,sensitivity,and F1 score of 0.951,0.950,0.990,0.950,and 0.950,respectively,without the need to specify lesion characteristics in advance,demonstrating its superiority in evaluation indicators.

Objective To evaluate the performance of an improved DeepSurv model for survival analysis in lung cancer patients,and investigate key factors affecting the prognosis of lung cancer.Methods The lung cancer data from the SEER database(2018-2021)was used in the study,and the DeepSurv model was optimized by incorporating a self-attention mechanism,a residual network,a LIME module and an entropy regularization term to enhance prediction accuracy and interpretability.Model performance was assessed using C-index and Brier score,and the improved model was utilized to evaluate the effects of various features on the prognosis of lung cancer.Results The improved DeepSurv model achieved a C-index of 0.852 and a Brier score of 0.139.Feature importance analysis identified age as the primary determinant of the survival of lung cancer patients.Conclusion The improved DeepSurv model outperforms both the Cox proportional hazards model and the original DeepSurv model in terms of accuracy,robustness and interpretability,offering a novel methodology for personalized medicine and survival analysis.

A converged network based on deep learning is proposed to realize the efficient and accurate diagnosis of diabetic retinopathy.Both data augmentation technology and generative adversarial network are used to augment the fundus images in EyePACS dataset for effectively addressing the problem of uneven classification of fundus images.The proposed model uses Inception-Resnet-V2 as the main network,and incorporates deep residual shrinkage network and pyramid split attention module for effectively filtering out the irrelevant information in the feature learning process and focusing on the lesion information,thereby improving the network's ability to capture important features.Experimental results show that the optimized model achieves accuracy,recall,specificity,sensitivity,and F1 score of 0.951,0.950,0.990,0.950,and 0.950,respectively,without the need to specify lesion characteristics in advance,demonstrating its superiority in evaluation indicators.

Objective To evaluate the performance of an improved DeepSurv model for survival analysis in lung cancer patients,and investigate key factors affecting the prognosis of lung cancer.Methods The lung cancer data from the SEER database(2018-2021)was used in the study,and the DeepSurv model was optimized by incorporating a self-attention mechanism,a residual network,a LIME module and an entropy regularization term to enhance prediction accuracy and interpretability.Model performance was assessed using C-index and Brier score,and the improved model was utilized to evaluate the effects of various features on the prognosis of lung cancer.Results The improved DeepSurv model achieved a C-index of 0.852 and a Brier score of 0.139.Feature importance analysis identified age as the primary determinant of the survival of lung cancer patients.Conclusion The improved DeepSurv model outperforms both the Cox proportional hazards model and the original DeepSurv model in terms of accuracy,robustness and interpretability,offering a novel methodology for personalized medicine and survival analysis.

Objective:To explore the feasibility and clinical efficacy of prone transpsoas lateral interbody fusion (PTP LIF) combined with posterior pedicle screw fixation for the treatment of lumbar degenerative diseases in the prone position.Methods:A total of 23 patients who underwent LLIF in the prone position at the First People's Hospital of Yunnan Province between March 2023 and October 2023 were retrospectively analyzed. The cohort comprised 9 males and 14 females, with a mean age of 55.5±8.8 years (range, 41-70 years). The clinical diagnoses included intervertebral disc herniation with endplate inflammation (3 cases), lumbar spinal stenosis (13 cases), lumbar spondylolisthesis (5 cases), and lumbar instability (2 cases). The surgical segments involved L 3, 4 (15 cases), L 4, 5 (6 cases), and L 3-L 5 (2 cases), with 21 cases involving a single segment and 2 cases involving double segments. The disc height and lumbar lordosis Angle before and after surgery were compared. Lower back pain was evaluated using the visual analogue scale (VAS), while lumbar spine function was assessed via the Oswestry Disability Index (ODI). Clinical efficacy was evaluated according to the modified MacNab criteria at the last follow-up. Results:All surgeries were successfully completed. The operation time was 120.2±21.4 min (range, 90-175 min), intraoperative blood loss was 131.1±40.8 ml (range, 60-200 ml), and the hospital stay was 6.2±1.6 days (range, 4-10 days). Follow-up was obtained for all 23 cases, with the follow-up time being 9.6±2.2 months (range, 6-13 months). One case of endplate damage occurred during surgery, two cases of transient psoas muscle weakness occurred postoperatively, and one case of lower limb pain and numbness was reported; no cases of wound infection or delayed healing were observed. The postoperative disc height improved compared to preoperative (6.8±1.9 mm; F=66.618, P<0.001). There was no statistically significant difference between 3 months postoperative (11.1±1.2 mm) and immediately postoperative (12.2±1.2 mm; P>0.05), but there was a statistically significant difference between the last follow-up (10.7±1.1 mm) and immediately postoperative ( P<0.05). The postoperative lumbar lordosis angle improved compared to preoperative (35.3°±5.4°; F=19.465, P<0.001), with no statistically significant difference between 3 months postoperative (44.1°±5.4°) and immediately postoperative (47.8°±6.6°; P>0.05), but there was a statistically significant difference between the last follow-up (43.2°±5.3°) and immediately postoperative ( P<0.05). The postoperative VAS score improved compared to preoperative (6.3±1.1 points; F=79.931, P<0.001), and the last follow-up (1.1±1.1 points) showed further improvement compared to 3 months postoperative (1.7±1.4 points; P<0.05). The postoperative ODI improved compared to preoperative (69.9%±7.4%; F=592.392, P<0.001), with 3 months postoperative (23.1%±3.1%) showing improvement compared to 1 month postoperative (29.2%±3.1%), and the last follow-up (17.5%±3.6%) showing further improvement compared to 3 months postoperative ( P<0.05). At the last follow-up, the modified MacNab criteria were: excellent in 16 cases, good in 5, fair in 2, with an excellent and good rate of 91% (21/23); 7 cases of cage subsidence were observed, with no cases of internal fixation loosening. Conclusion:PTP LIF combined with pedicle screw fixation for the treatment of lumbar degenerative diseases is safe and effective, with satisfactory short-term postoperative outcomes.

Objective:To analyze the distribution of clopidogrel metabolism-related gene variability in Kawasaki disease (KD) children with coronary artery lesions (CAL) across different age groups and the impact of genetic variability on the efficacy of clopidogrel antiplatelet therapy.Methods:A retrospective cohort study was conducted. Clinical data were collected from 46 KD children with CAL who were hospitalized in the Cardiovascular Center of Children′s Hospital of Fudan University between January 2021 and August 2022 and were treated with clopidogrel, including gender, age, body mass index, course of KD, CAL severity grade, and baseline platelet count. According to their age, the children were divided into ≥2-year-old group and <2-year-old group. Their platelet responsiveness was assessed by adenosine diphosphate-induced platelet inhibition rate (ADPi) calculated via thromboelastography, and children were categorized into high on-treatment platelet reactivity (HTPR) and normal on-treatment platelet reactivity (NTPR) groups. Genotypes of CYP2C19, PON1 and ABCB1 were detected. The t test, one-way analysis of variance and Chi-square test were used for intergroup comparison. Results:Among the 46 KD children with CAL, 34 were male and 12 were female; 37 were ≥2-year-old and 9 were <2-year-old; 25 cases were in the HTPR group and 21 cases were in the NTPR group, with 19 HTPR and 18 NTPR in the ≥2-year-old group, and 6 HTPR and 3 NTPR in the <2-year-old group. Genetic analysis showed that 92 alleles among the 46 children, with frequencies of CYP2C19*1, CYP2C19*2, CYP2C19*3, CYP2C19*17, PON1 192Q, PON1 192R, ABCB1 3435C, ABCB1 3435T at 59% (54/92), 32% (29/92), 9% (8/92), 1% (1/92), 36% (36/92), 64% (59/92), 63% (58/92) and 37% (34/92), respectively. Analysis of the impact of genotype on ADPi revealed that in children aged ≥2 years, those with CYP2C19*1/*3 genotype had significantly lower ADPi than those with CYP2C19*1/*1 genotype ((34±15)% vs. (61±29)%, t=2.18, P=0.036). There were also no significant difference in ADPi among children with PON1 192Q homozygous, PON1 192R heterozygote and PON1 192R homozygous genotypes ((40±22)% vs. (52±33)% vs. (65±27)%, F=2.17, P=0.130), or among those with ABCB1 3435C homozygous, ABCB1 3435T heterozygote and ABCB1 3435T homozygous genotypes ((55±34)% vs. (60±27)% vs. (49±24)%, F=0.33, P=0.719). In <2-year-old group, there were no significant differences in ADPi across CYP2C19*1/*1, CYP2C19*1/*2 and CYP2C19*2*2 genotypes ((40±20)% vs. (53±37)% vs. (34±16)%, F=0.37, P>0.05). There were no significant differences in ADPi across CYP2C19*1/*1 and CYP2C19*1/*3 genotypes ((44±27)% vs. (42±20)%, t=0.08, P>0.05). There were no significant differences in ADPi across PON1 192Q homozygous, PON1 192R heterozygote and PON1 192R homozygous genotypes (45% vs. (55±27)% vs. (24±5)%, F=1.83, P>0.05). There were no significant differences in ADPi across ABCB1 3435C homozygous, ABCB1 3435T heterozygote and ABCB1 3435T homozygous genotypes ((36±16)% vs. (50±35)% vs. 45%, F=0.29, P>0.05). The risk analysis of HTPR in different genotypes revealed that in children aged ≥2 years, carrying at least 1 or 2 loss-of-function alleles of CYP2C19 was a risk factor for HTPR ( OR=4.69, 10.00, 95% CI 1.11-19.83, 0.84-119.32, P=0.033, 0.046, respectively), and PON1 192R homozygosity and carrying at least one PON1 192R allele were protective factors against HTPR ( OR=0.08, 0.13, 95% CI 0.01-0.86, 0.01-1.19, P=0.019, 0.043, respectively). Conclusion:KD children aged ≥2 years carrying CYP2C19 loss-of-function alleles and PON1 192Q are more likely to develop HTPR.

Objective:To improve the ability of aviation physicians and physical examiners in differential diagnosis of primary angiitis of the central nervous system (PACNS) and the level of aeromedical assessment by exploring the diagnosis and treatment of PACNS.Methods:The clinical diagnosis and treatment of a pilot with PACNS were reported and the related literatures were reviewed.Results:The clinical manifestation of the fighter trainer pilot was numbness of the left finger, and then gradually appeared numbness and weakness of the left face and left upper limb more than 1 month. The initial diagnosis was viral encephalitis and acute disseminated encephalomyelitis and the hormonal shock treatment was ineffective. The final diagnosis was PACNS. The patient′s condition was improved by cyclophosphamide shock treatment. Aeromedical assessment concluded that the pilot was temporarily grounded.Conclusions:Pilot with PACNS is rarely reported and easy to be misdiagnosed due to the clinical and imaging findings are lack of specificity. The aeromedical assessment of PACNS pilots should be individualized according to the severity of illness, type of aircraft and flight duties.

Objective:To improve the ability of aviation physicians and physical examiners in differential diagnosis of primary angiitis of the central nervous system (PACNS) and the level of aeromedical assessment by exploring the diagnosis and treatment of PACNS.Methods:The clinical diagnosis and treatment of a pilot with PACNS were reported and the related literatures were reviewed.Results:The clinical manifestation of the fighter trainer pilot was numbness of the left finger, and then gradually appeared numbness and weakness of the left face and left upper limb more than 1 month. The initial diagnosis was viral encephalitis and acute disseminated encephalomyelitis and the hormonal shock treatment was ineffective. The final diagnosis was PACNS. The patient′s condition was improved by cyclophosphamide shock treatment. Aeromedical assessment concluded that the pilot was temporarily grounded.Conclusions:Pilot with PACNS is rarely reported and easy to be misdiagnosed due to the clinical and imaging findings are lack of specificity. The aeromedical assessment of PACNS pilots should be individualized according to the severity of illness, type of aircraft and flight duties.

A fast method composed of the quick, easy, cheap, effective, rugged and safe ( QuEChERS) and temperature-assisted ionic liquid dispersive liquid-liquid microextraction ( TA-IL-DLLME) sample preparation coupled with high performance liquid chromatography ( HPLC ) for the analysis of 5 dyes residues in navel orange was developed. The QuEChERS sample preparation involved the quick extraction with acetonitrile in the presence of anhydrous MgSO4 and NaCl and the purification with primary secondary amine ( PSA ) sorbent. The TA-IL-DLLME sample preparation was processed using 1 mL of the extract obtained by QuEChERS as dispersive solvent and 60 μL of 1-octyl-3-methylimidazolium hexafluorophosphate as extractive solvent under 55 ℃ of water-bath temperature and 12 min of water-bath time. The ultimate solution was detected by HPLC-UV and the contaminated sample was further confirmed by UPLC-MS/MS under multiple reactions monitoring (MRM) mode. The recoveries of five dyes were in the range from 70. 3% to 93. 6% at two spike levels of 0. 01 and 0. 05 mg/kg, the relative standard deviations (RSDs) were between 3. 5% and 9. 2% and the limits of quantification (LOQs) were between 1. 1 and 2. 8 μg/kg.

Objective To investigate the clinical features and prognosis of rectal neuroendocrine neoplasms (NENs) in different pathological grades.Methods The clinical data of 183 patients with rectal NENs who were admitted to the PLA General Hospital from January 2001 to April 2012 were retrospectively analyzed.All the clinical and pathological data of the patients who received endoscopy and (or) surgical resection were retrieved from the work station and the database of the endoscopic center.Based on the 2010 WHO pathology classification of digestive tumors,the pathological data were ranked according to the mitotic count.The prognosis of the patients was learned by re-examination or phone call.The follow-up ended till July 2014 or at the death of patients.Data were analyzed using the chi-square test.Results A total of 183 patients were enrolled in this study including 120 males and 63 females.The median age of the patients was 48 years (range,14-83 years).Seventy-four patients had the symptom of hemafecia,9 patients had abdominal pain and change in bowel habit,and 70 patients were diagnosed by body examination.Other symptoms included increased level of tumor markers and abdominal distension.Carcinoid syndrome was not detected in all the patients.Fourteen patients were complicated with polyp of intestine,5 with tubular adenoma,3 with colorectal adenocarcinoma and 1 with small cell lung cancer.The diameters of the tumor under 1 cm were detected in 162 patients,the diameters of the tumors ranged between 1 cm and 2 cm in 14 patients,and the diameters of the tumors above 2 cm in 7 patients.The mean distance between the tumor and the anus was (5-± 3)cm.Of the 183 patients,130 received endoscopic treatment,43 received surgical treatment and 10 received clamping because the tumor was misdiagnosed as polyps.There were 158 patients in grade1 (154 in stage Ⅰ,1 in stage Ⅱ,1 in stage Ⅲ and 2 in stage Ⅳ),21 in grade2 (13 in stage Ⅰ,3in stage Ⅱ,3 in stage Ⅲ and 2 in stage Ⅳ),4 in grade 3 (1 in stage Ⅰ,1 in stage Ⅲ and 2 in stage Ⅳ).Six patients had liver metastasis and 9 had lymph node metastasis.Fourteen patients died (4 in grade 1,6 in grade 2 and 4 in grade 3).The 5-year survival rate of patients was 92.35％ (169/183).There were significant differences in the gender,tumor diameter,tumor staging,lymph node metastasis distal metastasis and 5-year survival rate among patients with rectal NENs in different pathological grades (x2=60.949,71.587,32.135,55.486,56.512,P ＜ 0.05).Conclusions Rectal NENs lacks the specific clinical manifestation and is more likely to happen in males,and it often locates at the middle-lower rectum.Most of the rectal NENs belongs to stage Ⅰ and grade 1 and is less than 1 cm in size.The prognosis of patients with rectal NENs in different pathological grades is different.The 2010 WHO pathology classification of digestive tumors is useful to asses the prognosis of rectal NENs.Different grades of rectal NENs could be taken into account when designing the treatment plan.