With the rapid aging of the global population, the demand for prosthodontic rehabilitation of partial edentulism continues to increase. Among the available treatment options, removable partial dentures remain an essential clinical modality for restoring partially edentulous arches and masticatory function because of their broad indications, minimal invasiveness, and cost-effectiveness. However, unlike fixed prosthodontics, which have undergone substantial digitalization, removable partial denture therapy is characterized by dual tooth-tissue support, complex biomechanics, and highly variable anatomical morphology. As a result, the digital design and fabrication of removable partial dentures have long faced major bottlenecks, including limited standardization, steep learning curves, and heavy dependence on expert experience. In recent years, artificial intelligence (AI) has undergone a paradigm shift from early expert systems to machine learning and deep learning, creating new opportunities to address the complexity and precision demands of removable partial denture design. Based on a comprehensive literature review and our team's clinical experience, this article systematically elaborates on the specific applications and normative standards of AI throughout the entire removable partial denture fabrication workflow. In the data acquisition phase, we emphasize multimodal data fusion and digital pressure impression strategies. In the intelligent diagnosis and planning phase, deep learning is utilized to achieve the automatic identification of edentulous areas and undercuts, while machine learning and expert systems assist in abutment selection and treatment plan generation. Regarding automated design logic, the article analyzes the process from the intelligent planning of the common path of insertion to parametric design based on biomechanical optimization. At the clinical operation level, a standardized human-machine collaborative workflow of “AI generation-clinician review-local fine-tuning” is proposed. In terms of ethics and accountability, the principles of algorithmic transparency and “the clinician as the ultimate responsible subject” are emphasized. Furthermore, integrating clinical practice, this article innovatively proposes a “capability grading system for AI-assisted removable partial denture design (comprising the auxiliary analysis level, the rule-based design level, and the intelligent adaptive level)”. The aim is to offer practical recommendations for dental clinicians, dental technicians, and researchers, thereby facilitating the transition of removable partial denture prosthodontics toward greater precision, intelligence, and standardization.

ObjectiveTo quantitatively analyze and evaluate the content of China's assistive technology industry policies. MethodsBased on the three-dimensional framework of policy "objectives-instruments-intensity", this study examined 101 national-level assistive technology industry policies issued between 2011 and 2024. Policy text analysis was conducted to assess the matching and coordination among policy objectives, policy instruments and policy intensity. Furthermore, the policy modeling consistency (PMC) index model was applied to establish an evaluation system consisting of nine primary indicators and 35 secondary indicators, and six key policies were quantitatively evaluated. ResultsThe policy objectives mainly focused on optimizing the industrial development structure, also emphasized supportive environments for industrial development and the enhancement of innovation capacity. Policy instruments were dominated by supply-oriented and environment-oriented instruments, supplemented by demand-oriented instruments, aiming to strengthen technological foundations and competitiveness, promote standardized development, and facilitate industrial development and upgrading. Regarding the matching between policy objectives and instruments, multiple instruments were adopted to support different objectives, with different priorities. In terms of the matching between policy objectives and policy intensity, policies mainly emphasized improving the accessibility of assistive technology products and meeting the rehabilitation needs of different populations. Regarding the matching between policy instruments and policy intensity, supply-oriented instruments were more frequently adopted and showed higher policy intensity, indicating strong governmental efforts to optimize production factor allocation and enhance the capacity of market entities. The average PMC index score was 6.45, indicating relatively high policy quality, strong internal consistency, stable policy orientation, and clear objectives. The PMC scores for policy evaluation and policy instrument dimensions were relatively high, while some policies showed relatively narrow objectives. ConclusionChina's assistive technology industry policies demonstrate relatively high quality. Future policy development should further strengthen industrial innovation capacity, optimize the structure of policy instruments, and improve the alignment between policy objectives and instruments.

Objective:To analyze the efficacy of enzyme replacement therapy and anti-drug antibody production in children with Fabry disease.Methods:The clinical data of 7 children with Fabry disease treated with enzyme replacement therapy for more than 1 year at Children′s Hospital of Zhejiang University School of Medicine from July 2021 to June 2024 were retrospectively analyzed. The basic information and the changes of related clinical indicators before and after treatment were collected. Paired sample t test was used to compare renal function, left heart mass index, pain score and other related indexes before and after treatment. The anti-drug antibodies were detected by enzyme-linked immunosorbent assay. Results:A total of 6 boys and 1 girl were included. The age of diagnosis was (12.2±1.8) years. After 1 year of enzyme replacement therapy, the abnormal substrate globotriaosylsphingosine and brief pain inventory scores of all children were significantly lower than those before treatment ((16±11) vs. (63±42) μg/L, 22±19 vs. 45±29, t=3.88, 3.43, both P<0.05). There were no significant differences in glomerular filtration rate, urinary microalbumin to creatinine and left heart mass index before and after treatment ((124±35) vs. (136±26) ml/(min·1.73 m 2), (9.3±8.3) vs. (3.8±2.5) mg/g, (38±9) vs. (33±6) g/m 2.7, t=1.33, 1.74, 1.19, all P>0.05). Patients 4, 5 and 6 developed anti-drug antibodies at 1 month, 4 months and 1 month after medication, respectively. Patient 4 had persistently high anti-drug antibody titers (absorbance 3.65-3.73) accompanied by urticaria, elevated globotriaosylsphingosine and worsening clinical symptoms. Conclusions:The enzyme replacement therapy can effectively improve the clinical symptoms and reduce the level of globotriaosylsphingosine in children with Fabry disease. The anti-drug antibody is common in patients after long-term enzyme replacement therapy and may diminish the efficacy, which needs dynamic monitoring.

The paper reports a case of Fabry disease in a child with non-singular nocturnal enuresis as the first symptom. The boy developed unexplained nocturnal enuresis with frequent daytime urination since the age of 6. Fabry disease was detected and diagnosed by chance through high-risk screening. The activity of α-galactosidase A by dry blood spot was 1.97 μmol·L -1·h -1 , and there was c.640-801G>A mutation in GLA gene. Urine routine, urinary microprotein and renal function were normal. However, there were mulberry bodies found in urine deposition microscopy, suggesting the presence of kidney injury. This case suggests that enuresis can be the first symptom of Fabry disease, and mulberry bodies can be seen in the urine at the early stage of the disease.

Objective:To develop CNN-Dinov2, a deep learning-based automatic classification model for Gram-stained images, enabling rapid diagnosis of three prevalent Gram-negative bacilli in bloodstream infections: Escherichia coli ( E.coli), Klebsiella pneumoniae ( K.pneumoniae), and Pseudomonas aeruginosa ( P.aeruginosa). Methods:This evaluation study analyzed 1 425 Gram-stained microscopic images from patients with bloodstream infections at Houjie Hospital, in Dongguan City, collected between January 2023 and January 2024. The images, all positive for blood culture and identified as target strains, were categorized into Escherichia coli (419 images), Klebsiella pneumoniae (411 images), Pseudomonas aeruginosa (413 images), and other Gram-negative bacilli (182 images). They were randomly split into a training set (1 141 images), a validation set (141 images), and a test set (143 images) in an 8∶1∶1 ratio. A hybrid CNN-Dinov2 model was developed by integrating ResNet′s local feature extraction with Dinov2′s global pre-trained features, followed by a fully connected layer. The model was optimized by inputting the preprocessed images and adjusting parameters through loss calculation and backpropagation. AlexNet, Dinov2, and ResNet18 served as control models. The models′ classification performance was assessed using accuracy, precision, weighted F1 score, and recall rate, derived from the confusion matrix. The PR curve and AP value further evaluated each model′s classification capability across the four image categories. Results:The CNN-Dinov2 model achieved a training accuracy of 99.74%, a validation accuracy of 98.12%, and a validation loss of 0.070 6, demonstrating robust generalization without overfitting. Validation metrics revealed superior performance with an accuracy of 98.60%, precision of 98.65%, a weighted F1 score of 98.60%, and a recall rate of 98.60%, outperforming other models. The confusion matrix confirmed its strong classification capability, with the highest sum of diagonal values for identifying four types of bacteria. The macro average precision (AP) values under the precision-recall (PR) curves were all 1, indicating excellent discrimination across all categories. Overall, the CNN-Dinov2 model exhibited the best performance among the four models evaluated.Conclusion:This study successfully developed CNN-Dinov2, an automated classification model for Gram staining images. It offers valuable support for the rapid diagnosis of bloodstream infections caused by Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, demonstrating practical utility.

Digital technologies have become an integral part of complete denture restoration. With advancement in computer-aided design and computer-aided manufacturing (CAD/CAM), tools such as intraoral scanning, facial scanning, 3D printing, and numerical control machining are reshaping the workflow of complete denture restoration. Unlike conventional methods that rely heavily on clinical experience and manual techniques, digital technologies offer greater precision, predictability, and efficacy. They also streamline the process by reducing the number of patient visits and improving overall comfort. Despite these improvements, the clinical application of digital complete denture restoration still faces challenges that require further standardization. The major issues include appropriate case selection, establishing consistent digital workflows, and evaluating long-term outcomes. To address these challenges and provide clinical guidance for practitioners, this expert consensus outlines the principles, advantages, and limitations of digital complete denture technology. The aim of this review was to offer practical recommendations on indications, clinical procedures and precautions, evaluation metrics, and outcome assessment to support digital restoration of complete denture in clinical practice.
Humans ; Denture, Complete ; Computer-Aided Design ; Denture Design/methods* ; Consensus ; Printing, Three-Dimensional

The paper reports a case of Fabry disease in a child with non-singular nocturnal enuresis as the first symptom. The boy developed unexplained nocturnal enuresis with frequent daytime urination since the age of 6. Fabry disease was detected and diagnosed by chance through high-risk screening. The activity of α-galactosidase A by dry blood spot was 1.97 μmol·L -1·h -1 , and there was c.640-801G>A mutation in GLA gene. Urine routine, urinary microprotein and renal function were normal. However, there were mulberry bodies found in urine deposition microscopy, suggesting the presence of kidney injury. This case suggests that enuresis can be the first symptom of Fabry disease, and mulberry bodies can be seen in the urine at the early stage of the disease.

Objective:To develop CNN-Dinov2, a deep learning-based automatic classification model for Gram-stained images, enabling rapid diagnosis of three prevalent Gram-negative bacilli in bloodstream infections: Escherichia coli ( E.coli), Klebsiella pneumoniae ( K.pneumoniae), and Pseudomonas aeruginosa ( P.aeruginosa). Methods:This evaluation study analyzed 1 425 Gram-stained microscopic images from patients with bloodstream infections at Houjie Hospital, in Dongguan City, collected between January 2023 and January 2024. The images, all positive for blood culture and identified as target strains, were categorized into Escherichia coli (419 images), Klebsiella pneumoniae (411 images), Pseudomonas aeruginosa (413 images), and other Gram-negative bacilli (182 images). They were randomly split into a training set (1 141 images), a validation set (141 images), and a test set (143 images) in an 8∶1∶1 ratio. A hybrid CNN-Dinov2 model was developed by integrating ResNet′s local feature extraction with Dinov2′s global pre-trained features, followed by a fully connected layer. The model was optimized by inputting the preprocessed images and adjusting parameters through loss calculation and backpropagation. AlexNet, Dinov2, and ResNet18 served as control models. The models′ classification performance was assessed using accuracy, precision, weighted F1 score, and recall rate, derived from the confusion matrix. The PR curve and AP value further evaluated each model′s classification capability across the four image categories. Results:The CNN-Dinov2 model achieved a training accuracy of 99.74%, a validation accuracy of 98.12%, and a validation loss of 0.070 6, demonstrating robust generalization without overfitting. Validation metrics revealed superior performance with an accuracy of 98.60%, precision of 98.65%, a weighted F1 score of 98.60%, and a recall rate of 98.60%, outperforming other models. The confusion matrix confirmed its strong classification capability, with the highest sum of diagonal values for identifying four types of bacteria. The macro average precision (AP) values under the precision-recall (PR) curves were all 1, indicating excellent discrimination across all categories. Overall, the CNN-Dinov2 model exhibited the best performance among the four models evaluated.Conclusion:This study successfully developed CNN-Dinov2, an automated classification model for Gram staining images. It offers valuable support for the rapid diagnosis of bloodstream infections caused by Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, demonstrating practical utility.

Objective To investigate the effects of transcutaneous electrical acupoint stimulation at Neiguan combined with Danzhong on intraoperative remifentanil dosage and early postoperative pain in pa-tients undergoing radical mastectomy.Methods A total of 153 female patients undergoing radical mastecto-my under general anaesthesia aged 18-64 years,BMI 18.5-30.0 kg/m2 and ASA physical status Ⅰ or Ⅱwere enrolled.The patients were randomly assigned into three groups:single-acupoint(Neiguan)group,dual-acupoint(Neiguan and Danzhong)group and control group,51 patients in each group.TEAS for 30 minutes at different points(Neiguan combined with Danzhong for the dual-acupoint group,Neiguan for the single-acupoint group,no stimulation for the sham group)were given before anaesthesia induction.30 mi-nutes of disperse/dense stimulation of 2/10 Hz with an intensity of 4-11 mA was used.The consumption of remifentanil,time to recall and to extubation were recorded.The visual analogue scale(VAS)pain score in PACU and 24 hours after surgery,and patient satisfaction scores was recorded.Results The intra-operative remifentanil consumption in the dual-acupoint group was significantly less than groups single-acupoint and sham(P＜0.05).Patient satisfaction score 24 hours after surgery was significantly higher in the dual-acu-point group than those in groups single-acupoint and sham(P＜0.05).The VAS scores in PACU and 24 hours after surgery had no significantly differences among the three groups.Conclusion Compared with single-acupoint,stimulation at Neiguan combined with Danzhong in patients undergoing radical mastectomy induced stronger remifentanil-sparing effect and higher patients'satisfaction after radical mastectomy.

Objective:To explore the clinical and genetic characteristics of a child with mental retardation, language and motor developmental delay and epilepsy.Methods:A Child who was admitted to the First Affiliated Hospital of Zhengzhou University in March 2020 for intermittent seizures for over two months was Selected as the study subject. Clinical data of the child was collected. Peripheral blood samples of the child and his parents were collected and subjected to high throughput sequencing. Candidate variants were verified by Sanger sequencing and bioinformatic analysis.Results:The clinical manifestations of the child have included mental retardation, language and motor developmental delay, and seizures. High-throughput sequencing revealed that he has harbored a hemizygous splice site variant (NM_032591.3: c.1030-1G>C) of the SLC9A7 gene, which was inherited from his mother and unreported previously. Conclusion:The hemizygous splice site variant (NM_032591.3: c. 1030-1G>C) of the SLC9A7 gene probably underlay the disease in this child. Above finding has provided a basis for clinical diagnosis and genetic counseling.