ObjectiveTo establish three standardized rehabilitation unit models covering acute, critical and comprehensive recovery stages for traumatic brain injury (TBI) that fit the whole-course characteristics, clarify their multidisciplinary collaborative framework, standardized procedures, and information management system, and verify the clinical effectiveness of this staged rehabilitation unit on long-term functional prognosis and in-hospital complications of TBI patients. MethodsThis study was conducted at two stages: model construction and efficacy verification. At the first stage, TBI rehabilitation units for acute, critical and recovery stages were established in the Department of Neurosurgery, Intensive Care Unit and Department of Rehabilitation Medicine, respectively, with unified multidisciplinary team configuration, standardized assessment and treatment protocols, collaborative workflows, and a dedicated information management system. At the second stage, 324 TBI patients admitted to the newly built rehabilitation units from July, 2023 to July, 2025 were enrolled as the observation group, and 167 TBI patients receiving traditional management from January, 2018 to December, 2022 were retrospectively selected as the control group. The primary outcomes were 1-year post-injury Glasgow Outcome Scale (GOS) and Disability Rating Scale (DRS) scores. Secondary outcomes included the incidence of in-hospital complications and the mean indwelling time of tracheal intubation, gastric tube and urinary catheter. ResultsThe 1-year GOS score was significantly higher (t = -7.836, P < 0.001), and the DRS score was significantly lower (t = -6.742, P < 0.001) in the observation group than in the control group. The overall incidence of complications was significantly lower in the observation group than in the control group (χ² = 18.762, P < 0.001). The incidences of pulmonary infection, urinary tract infection, pressure ulcer, lower extremity venous thrombosis and joint contracture deformity were all significantly reduced (χ² > 4.328, P < 0.05). The indwelling time of tracheal intubation, gastric tube and urinary catheter was significantly shorter in the observation group than in the control group (t > 10.276, P < 0.001). ConclusionThe acute-critical-comprehensive three-stage standardized TBI rehabilitation unit could significantly improve the long-term functional outcomes of TBI patients, reduce the incidence of complications, and shorten the indwelling time of invasive tubes through multidisciplinary collaboration, whole-course standardized management and information-based quality control. This model conforms to the pathological and functional laws of TBI, realizes seamless whole-course connection from acute treatment to recovery rehabilitation, and has good clinical feasibility and promotion value. It can provide evidence-based evidence and practical reference for the homogenized and standardized development of TBI rehabilitation in China.

Objective:To establish an automatic classification approach for bone marrow cells based on microscopic hyperspectral imaging and three-dimensional spectral convolutional neural network (Spec-CNN).Methods:The research type is establishment of methodology. The study included 306 newly diagnosed patients' bone marrow smears under Wright's staining from the Department of Hematology of the First Medical Center of the PLA General Hospital from November 1st, 2013 to April 30th, 2024. The high-spectrum data and 4k image data of bone marrow cells were simultaneously collected using a microscopic hyperspectral-4k optical path integrated imaging system (with a spectral resolution of 400—1 000 nm). The high-spectrum data was used for model training, while the 4k image data recognized by morphologists was only used as a reference for labeling the high-spectrum data. The high-spectrum data set was divided into training set, validation set and test set in a ratio of 14∶6∶5. The training set and validation set were used to train and fine-tune the Spec-CNN model, and the test set was used to evaluate the model performance. The sensitivity, specificity ,accuracy ,and Kappa coefficient were calculated for comparing the manual annotation results as gold standard with the intelligent identification results of the Spec-CNN model. Five non-data set samples were used for external validation.Results:The acquired hyperspectral data and 4k imaging dataset comprised of 32 categories and 64 800 bone marrow cells. In the test set, the Spec-CNN model demonstrated weighted-average indicators on classification metrics across 32 cell types: sensitivity 87.79%, specificity 99.31%, and accuracy 98.78%, and Kappa coefficient 0.869. For external validation, the mean correct identification rate of bone marrow cells reached 83.28%.Conclusion:We successfully established an automatic classification method of bone marrow cells based on microscopic hyperspectral imaging and three-dimensional Spec-CNN. This method has a good automatic classification ability for 32 types of bone marrow nucleated cells, which has a certain auxiliary effect on improving the diagnosis efficiency of blood diseases for bone marrow morphologists.

Objective:To investigate the enhancement effect of deep progressive reconstruction (DPR) algorithm on image quality and standardized uptake value (SUV) of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) in obese patients. Methods:The images of a total of 27 obese patients,who underwent 18F-FDG PET/CT in Affiliated Nanjing Hospital of Nanjing Medical University from September 2023 to May 2024,were retrospectively enrolled. The images of all patients were collected by using uMI 780 PET/CT. Ordered subset expectation maximization (OSEM) iterative algorithm and DPR algorithm were adopted to reconstruct PET images,and measure counting rate of scattering coincidence of PET/CT images,counting rate of true coincidence,noise equivalent counting rate (NECR) and scattering fraction (SF). The quality indicators of PET images included signal-to-noise ratio (SNR),the maximum SUV (SUVmax) of lesions,the tumor-to-background ratio (TBR),the contrast-to-noise ratio (CNR) and the visual scores of 18F-FDG PET/CT images on livers were evaluated. The differences and consistency of various indicators between DPR and OSEM reconstruction algorithms were further analyzed. Results:The average 18F-FDG PET/CT injection activity of 27 patients was (0.12±0.01) mCi (1 mCi=37 MBq)/kg,and the counting rate of true coincidence,NECR and SF of PET images were respectively (153.73±25.09),(44.81±8.47) kcps and (36.77±1.91)％. The SNR of liver obtained by DPR algorithm was (15.83±3.60),which was significantly higher than that (9.06±1.87) of OSEM algorithm,with statistically significant (t=20.6,P<0.05),and there was significantly correlation in liver SNR between two algorithms (R2=0.91,P<0.05). In 27 uptake 18F-FDG PET/CT lesions,the SUVmax,TBR and CNR of lesions that were obtained from OSEM algorithm were respectively (5.86±1.49),(1.95±0.49) and (17.74±4.77),which were lower than corresponding those of DPR algorithm,and the differences were significant (t=9.03,8.79,15.49,P<0.05),respectively. There were significant correlations in SUVmax,TBR and CNR between the two algorithms (R2=0.71,0.70,0.76,P<0.05),respectively. The visual scores of PET images obtained from the DPR algorithm was 4 (3,5) scores,which was significantly higher than 3 (2,4) scores of OSEM algorithm,and the difference of that between two algorithms was significant (U=396,P<0.05). Conclusion:The scattering effect of 18F-FDG PET/CT imaging is stronger in obese patients,whose counting rate of equivalent effect of noise is lower. The DPR reconstruction algorithm can significantly improve the SNR and lesion contrast of PET images than the OSEM algorithm,which has significant gain effect on the SUVmax of lesions,and it can significantly improve the quality of 18F-FDG PET/PET images in obese patients.

Objective:To establish an automatic classification approach for bone marrow cells based on microscopic hyperspectral imaging and three-dimensional spectral convolutional neural network (Spec-CNN).Methods:The research type is establishment of methodology. The study included 306 newly diagnosed patients' bone marrow smears under Wright's staining from the Department of Hematology of the First Medical Center of the PLA General Hospital from November 1st, 2013 to April 30th, 2024. The high-spectrum data and 4k image data of bone marrow cells were simultaneously collected using a microscopic hyperspectral-4k optical path integrated imaging system (with a spectral resolution of 400—1 000 nm). The high-spectrum data was used for model training, while the 4k image data recognized by morphologists was only used as a reference for labeling the high-spectrum data. The high-spectrum data set was divided into training set, validation set and test set in a ratio of 14∶6∶5. The training set and validation set were used to train and fine-tune the Spec-CNN model, and the test set was used to evaluate the model performance. The sensitivity, specificity ,accuracy ,and Kappa coefficient were calculated for comparing the manual annotation results as gold standard with the intelligent identification results of the Spec-CNN model. Five non-data set samples were used for external validation.Results:The acquired hyperspectral data and 4k imaging dataset comprised of 32 categories and 64 800 bone marrow cells. In the test set, the Spec-CNN model demonstrated weighted-average indicators on classification metrics across 32 cell types: sensitivity 87.79%, specificity 99.31%, and accuracy 98.78%, and Kappa coefficient 0.869. For external validation, the mean correct identification rate of bone marrow cells reached 83.28%.Conclusion:We successfully established an automatic classification method of bone marrow cells based on microscopic hyperspectral imaging and three-dimensional Spec-CNN. This method has a good automatic classification ability for 32 types of bone marrow nucleated cells, which has a certain auxiliary effect on improving the diagnosis efficiency of blood diseases for bone marrow morphologists.

Objective:To investigate the enhancement effect of deep progressive reconstruction (DPR) algorithm on image quality and standardized uptake value (SUV) of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) in obese patients. Methods:The images of a total of 27 obese patients,who underwent 18F-FDG PET/CT in Affiliated Nanjing Hospital of Nanjing Medical University from September 2023 to May 2024,were retrospectively enrolled. The images of all patients were collected by using uMI 780 PET/CT. Ordered subset expectation maximization (OSEM) iterative algorithm and DPR algorithm were adopted to reconstruct PET images,and measure counting rate of scattering coincidence of PET/CT images,counting rate of true coincidence,noise equivalent counting rate (NECR) and scattering fraction (SF). The quality indicators of PET images included signal-to-noise ratio (SNR),the maximum SUV (SUVmax) of lesions,the tumor-to-background ratio (TBR),the contrast-to-noise ratio (CNR) and the visual scores of 18F-FDG PET/CT images on livers were evaluated. The differences and consistency of various indicators between DPR and OSEM reconstruction algorithms were further analyzed. Results:The average 18F-FDG PET/CT injection activity of 27 patients was (0.12±0.01) mCi (1 mCi=37 MBq)/kg,and the counting rate of true coincidence,NECR and SF of PET images were respectively (153.73±25.09),(44.81±8.47) kcps and (36.77±1.91)％. The SNR of liver obtained by DPR algorithm was (15.83±3.60),which was significantly higher than that (9.06±1.87) of OSEM algorithm,with statistically significant (t=20.6,P<0.05),and there was significantly correlation in liver SNR between two algorithms (R2=0.91,P<0.05). In 27 uptake 18F-FDG PET/CT lesions,the SUVmax,TBR and CNR of lesions that were obtained from OSEM algorithm were respectively (5.86±1.49),(1.95±0.49) and (17.74±4.77),which were lower than corresponding those of DPR algorithm,and the differences were significant (t=9.03,8.79,15.49,P<0.05),respectively. There were significant correlations in SUVmax,TBR and CNR between the two algorithms (R2=0.71,0.70,0.76,P<0.05),respectively. The visual scores of PET images obtained from the DPR algorithm was 4 (3,5) scores,which was significantly higher than 3 (2,4) scores of OSEM algorithm,and the difference of that between two algorithms was significant (U=396,P<0.05). Conclusion:The scattering effect of 18F-FDG PET/CT imaging is stronger in obese patients,whose counting rate of equivalent effect of noise is lower. The DPR reconstruction algorithm can significantly improve the SNR and lesion contrast of PET images than the OSEM algorithm,which has significant gain effect on the SUVmax of lesions,and it can significantly improve the quality of 18F-FDG PET/PET images in obese patients.

Nonobstructive azoospermia (NOA), one of the most severe types of male infertility, etiology often remains unclear in most cases. Therefore, this study aimed to detect four biallelic detrimental variants (0.5%) in the minichromosome maintenance domain containing 2 ( MCMDC2 ) genes in 768 NOA patients by whole-exome sequencing (WES). Hematoxylin and eosin (H&E) demonstrated that MCMDC2 deleterious variants caused meiotic arrest in three patients (c.1360G>T, c.1956G>T, and c.685C>T) and hypospermatogenesis in one patient (c.94G>T), as further confirmed through immunofluorescence (IF) staining. The single-cell RNA sequencing data indicated that MCMDC2 was substantially expressed during spermatogenesis. The variants were confirmed as deleterious and responsible for patient infertility through bioinformatics and in vitro experimental analyses. The results revealed four MCMDC2 variants related to NOA, which contributes to the current perception of the function of MCMDC2 in male fertility and presents new perspectives on the genetic etiology of NOA.
Humans ; Male ; Azoospermia/genetics* ; Meiosis/genetics* ; Spermatogenesis/genetics* ; Adult ; Exome Sequencing ; Microtubule-Associated Proteins/genetics* ; Alleles ; Infertility, Male/genetics*

OBJECTIVES: To study the impact of family socioeconomic status on children's reading fluency and the chain mediation effect of family reading environment and children's living and learning styles in this relationship. METHODS: A total of 473 children from grades 2 to 6 in two primary schools in Nanjing were selected through stratified random sampling. The children's reading fluency was assessed, and a questionnaire was used to collect information on family socioeconomic status, family reading environment, and children's living and learning styles. The mediation model was established using the Process macro in SPSS, and the Bootstrap method was employed to test the significance of the mediation effects. RESULTS: Family socioeconomic status, family reading environment, and children's living and learning styles were significantly positively correlated with reading fluency (P<0.001). The family reading environment and children's living and learning styles mediated the relationship between family socioeconomic status and children's reading fluency. Specifically, the independent mediation effect of family reading environment accounted for 11.02% of the total effect, while the independent mediation effect of children's living and learning styles accounted for 10.79%. The chain mediation effect of family reading environment and children's living and learning styles accounted for 7.41% of the total effect. CONCLUSIONS Family socioeconomic status can affect children's reading fluency through three pathways: family reading environment, children's living and learning styles, and the chain mediation effect of family reading environment and children's living and learning styles.
Humans ; Child ; Male ; Female ; Reading ; Learning ; Social Class ; Family

Objective:To explore the safety and efficacy of intraosseous regional administration (IORA) of vancomycin for preventing infection in primary total knee arthroplasty (TKA).Methods:A total of 124 patients with knee osteoarthritis undergoing TKA between February 2024 and May 2024 at nine hospitals were enrolled. Preoperative infection prophylaxis involved either IORA (0.5 g vancomycin administered via intraosseous regional infusion before incision) or intravenous infusion (1 g vancomycin via peripheral vein). The IORA group included 15 males and 47 females with a median age of 66.5 years (range, 60.0-70.0 years), while the intravenous group included 14 males and 48 females with a median age of 66.0 years (range, 61.8-70.3 years) years. Intraoperative samples were collected including fat and synovium tissues after incision, before prosthesis placement, and after tourniquet release; distal femoral cancellous bone during femoral osteotomy; proximal tibial cancellous bone during tibial osteotomy; proximal intercondylar cancellous bone before prosthesis placement; and peripheral blood from non-infused arms at surgery initiation and after tourniquet release. Vancomycin concentrations were measured using liquid chromatography-tandem mass spectrometry. Vital sign changes were recorded from admission to 5~10 minutes post-IORA (IORA group) or post-incision (intravenous group). Follow-ups were conducted on postoperative day 1 and 3, and at 1 and 3 months, to document complications including IORA-related adverse events, periprosthetic joint infections, surgical site infections, red man syndrome, acute kidney injury, deep vein thrombosis and so on.Results:Vancomycin concentrations in bone, fat, and synovial tissue samples were significantly higher in the IORA group than in the intravenous group ( P<0.05), while vancomycin concentrations in blood samples were significantly lower in the IORA group than in the intravenous group ( P<0.05). Only 7.3%(41/558) of tissue samples in the IORA group had vancomycin concentrations below 2.0 μg/g (the minimum inhibitory concentration of vancomycin against coagulase-negative staphylococcus), compared to 59.3%(331/558) in the intravenous group (χ 2=11.285, P<0.001). In the intravenous group, 16.9%(21/124) of blood samples had vancomycin concentrations exceeding 15.0 mg/L (the threshold associated with a significantly increased risk of nephrotoxicity), while all concentrations in the IORA group were below this threshold, the difference was statistically significant (χ 2=22.943, P<0.001). There were no statistically significant difference ( P>0.05) in vital signs changes before and after vancomycin administration between the two groups. Two patients in the intravenous group experienced incision exudate, while no other related complications occurred in either group. Conclusions:Compared to the traditional intravenous infusion of 1 g vancomycin, intraosseous injection of a low dose (0.5 g) of vancomycin achieves higher local tissue concentrations in the knee joint with a lower incidence of adverse reactions and is safe for infection prophylaxis. Despite guidelines not recommending the routine use of vancomycin for preventing infection after primary TKA, intraosseous injection of 0.5 g vancomycin may be considered intraoperatively for primary TKA in the following scenarios: patients in medical institutions with a high prevalence of methicillin-resistant staphylococcus aureus (MRSA) infections, patients with potential preoperative MRSA colonization, or patients with cephalosporin allergy.

Objective To analyze the results of ultrasonic examinations of hip joints in preterm infants,clarify the developmental characteristics of hip joints across different gestational ages.Meth-ods A total of 881 preterm infants who attended the Child Healthcare Outpatient Clinic of Nanjing Maternity and Child Health Care Hospital between January 2023 and April 2024 were enrolled.The first ultrasonic examination of the hip joints was performed at a corrected gestational age of 4 to 6 weeks.The Graf classification method was employed to categorize the ultrasonic findings.Differences in hip joint development among preterm infants of varying gestational ages were compared,and rele-vant influencing factors were statistically analyzed.Results A total of 1,762 hip joints were exam-ined,with 30 cases of abnormal hip joints detected,yielding an abnormality detection rate of 3.4％(30/881).Among these,26 hips were classified as type Ⅱa and 8 as type Ⅱb.Statistically signifi-cant differences were observed in the detection of left-sided abnormal hip joints between preterm in-fants of other gestational age groups and late preterm infants(P＜0.05).A statistically significant difference was found in the classification of right-sided hip joints between preterm infants of different genders(P＜0.05),while no significant difference was observed for left-sided hip joints(P＞0.05).Statistically significant differences were noted in the detection rates of bilateral abnormal hip joints among preterm infants with varying birth weights(P＜0.05).Advanced maternal age was found to influence the development of left-sided hip joints inpreterm infants,whereas factors such as assisted reproductive technology,breech presentation,and nulliparity had no significant impacts on abnormal hip joint development in preterm infants.Conclusion Ultrasonic technology demonstrates signifi-cant advantages in screening for developmental dysplasia of the hip(DDH)in preterm infants.Dif-ferent gestational ages exert a notable influence on the development of left-sided hip joints in preterm infants.Female preterm infants exhibit a higher risk of right-sided hip joint abnormalities compared to the left side.Breech presentation,assisted reproductive technology,and nulliparity are not risk factors for DDH in preterm infants.However,higher birth weight in preterm infants and advanced maternal age are associated with abnormal hip joint development.Clinical attention should focus on high-risk factors,and enhanced dynamic ultrasonic monitoring of hip joints in preterm infants is war-ranted to facilitate early intervention and treatment.

Objective To explore an assessment model for multi-level integrated training in health service based on advanced intelligent trauma simulators in order to innovate health service training.Methods An assessment model for multi-level integrated training that involved advanced trauma simulators was adopted to assess chains of treatment and rescue that were composed of multi-hierarchy medical institutions.The assessment focused on trauma emergency response capabilities at each level and the overall therapeutic effect.Results In terms of capabilities for treatment and rescue,group B was the best one,followed by group C and group A.As for the overall therapeutic effect,group A was outstanding,followed by group B and group C.Based on the priorities of combat casualty care,the final results of assessment were as follows:group A was the best one,followed by group B and group C.Conclusion The analysis of processes and outcomes of assessment reveals the edge of this model,as evidenced by the continuity of treatment and rescue,integrity of overall effectiveness,and adaptability of this assessment model.