Objective:To develop and evaluate a deep learning-based automatic measurement model for panoramic ultrasound biomicroscopy (UBM) images.Methods:A diagnostic test study was conducted.Preoperative UBM examination results of 372 patients who underwent implantable collamer lens (ICL) implantation were collected at the Eye Center of Renmin Hospital of Wuhan University between February 2021 and March 2023.A total of 1 368 panoramic UBM images were obtained to establish an image database.The dataset was divided into a training set (760 images), a validation set (86 images) and an internal test set (522 images).An expert panel consisting of three ophthalmologists annotated the images.The UNet+ + network was used to automatically segment anterior segment tissues, such as the cornea, lens and iris.In addition, image processing techniques and geometric localization algorithms were developed to automatically identify the anatomical landmarks of pupil diameter (PD), anterior chamber depth (ACD), angle-to-angle distance (ATA) and sulcus-to-sulcus distance (STS) to complete the measurement of these parameters.Additionally, 480 panoramic UBM images of 135 patients (240 eyes) from Aier Eye Hospital of Wuhan University were used as an external test set to further evaluate the performance of the model in different centers.The consistency between the measurements from the model and expert panel, the Pentacam system was assessed.Finally, 150 images were randomly selected from the external test set for a human-machine comparison to further evaluate the model's performance.This study adhered to the Declaration of Helsinki.The study protocol was approved by the Ethics Committee of Renmin Hospital of Wuhan University (No.WDRY-2022-K109) and Aier eye Hospital of Wuhan University (No.2023IRBKY120903).Written informed consent was obtained from each subject.Results:In the internal test dataset and external test dataset, with manual labeling as the reference standard, the model achieved a mean Dice coefficient of not less than 0.882.At least 95.65% of the anatomical landmark localization results had Euclidean distance differences within 250 μm.The intraclass correlation coefficients (ICCs) for the measurements of PD, ACD, angle-to-angle ATA, and STS were at least 0.958, with mean relative errors not exceeding 2.407%.With the Pentacam measurements as the reference standard, the ICCs for PD in the internal and external test sets were 0.540 and 0.466, respectively, while the ICCs for ACD were 0.946 and 0.908, respectively.In the human-machine comparison, the ICCs between the model's measurements and those of senior experts were all not lower than 0.969.Conclusions:The deep learning-based model can automatically measure anterior segment parameters from preoperative panoramic UBM images of patients undergoing ICL surgery.The model demonstrates a consistency comparable to that of senior experts, while providing higher efficiency.In terms of ACD measurement, the model shows good agreement between the measurements obtained from the model and Pentacam system.

Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

Objective To assess the clinical utility of single-molecule real-time technology(SMRT)in identifying triplicated α-globin genes and compound variant alleles.Methods A total of 36 samples with tripli-cated α-globin genes were collected.Among them,28 samples were confirmed by PCR flow-through hybridiza-tion and 8 samples were confirmed by Next Generation Sequencing(NGS).These 36 samples included tripli-cated α-globin genes compound variants with cis or trans arrangements unknown,such as αααanti4 2 compoundαcsα(2 cases),αααanti4.2 compound-α3.7(10 cases),and HKαα/--SEA pending confirmation(2 cases),SMRT technology was employed to detect thalassemia gene variants.Additionally,a pedigree with the genotype ofαααanti4.2 compound-α3.7 variant was recruited,including the proband(Ⅱ-1),its father(Ⅰ-1),and mother(Ⅰ-2).PCR flow-through hybridization and SMRT were employed to detect thalassemia gene variants.Results SMRT detected 35 out of 36 samples with triplicated α-globin genes,and 1 sample with quadrupllcated α-globin genes(ααααanti4.2).Among the 2 αααanti4 2 compound αCSα variant samples,both αααanti42 and αCSα were arranged in trans,with a genotype of αααanti4.2/αCSα.Among the 10 αααanti4.2 compound-α3.7 variant samples,9 samples hadαααanti4.2 and-α3.7 in a cis arrangement,with a genotype of HKαα/αα,and 1 sample had αααannti4.2 and-α3.7 in a trans arrangement,with a genotype of αααanti4.2/-α3.7.Compared with PCR flow-through hybridization,SMRT detected one case of a large segment deletion in the β-globin gene and two unknown variants,which led to an increase in the positive detection rate of approximately 10.71％(3/28).The pedigree analysis showed that the proband(Ⅱ-1)inherited αααanti4.2 and-α3.7 variants from his mother(Ⅰ-2),with a genotype of HKαα/αα,con-sistent with the SMRT detection results.Conclusion SMRT can accurately detect triplicated or quadrupllcat-ed α-globin genes,and compound variant alleles.It offers high accuracy,enables one-step identification of cis or trans arrangements,and provides comprehensive coverage of thalassemia gene variations,demonstrating its significant clinical value.

Objective:To evaluate potentially inappropriate medication (PIM) in hospitalized elderly patients with bacterial pneumonia, and explore its influencing factors.Methods:It was a single-center cross-sectional study. The study focused on elderly patients with bacterial pneumonia who were admitted to Xuanwu Hospital, Capital Medical University from January 2018 to November 2022. Patients′ gender, age, weight, length of hospital stay, diagnosis at admission, physical examination, diagnosis at discharge, comorbidities, medications, and laboratory test results were extracted from hospital information system and electronic medical records. Medication use of patients included in the analysis during their hospitalization were evaluated according to the classification of PIMs in the 5 lists of the Beer′s criteria of American Geriatrics Society. Based on whether PIM occurred, the patients were divided into with PIM group and without PIM group. The clinical features between the 2 groups were compared and the influencing factors of PIM were analyzed using multivariable logistic regression.Results:A total of 2 720 patients were included, in which 1 734 (63.75%) were male. The median age was 78 (70, 85) years and their ages ranged from 65 to 103 years. The number of drugs used per patient was 14 (10, 18) kinds, ranging from 1 to 57 kinds. The length of hospital stay was 12 (9, 17) days, ranging from 1 to 162 days. Charlson comorbidity index (CCI) was 6 (5, 8) points. Among the 2 720 patients, 1 894 (69.63%) experienced PIM, with a total of 6 166 cases of PIM. The top 3 drugs ranked by the number of PIM occurrence were antiplatelet agents (1 357 cases), benzodiazapine receptor agonists (956 cases), and antipsychotics (884 cases). The comparison of clinical characteristics between the 2 groups showed that differences in age, CCI, length of hospital stay, and number of medications between with PIM and without PIM patients were statistically significant (all P<0.001). Multivariable logistic regression results showed that CCI, length of hospital stay, and number of medications were independent influencing factors for PIM. The risk increased by 8% and 1% with one point increase in CCI and one day extension in length of hospital stay [odds ratio ( OR)=1.08, 95% confidence interval ( CI): 1.04-1.13, P<0.001; OR=1.01, 95% CI: 1.00-1.03, P=0.03]. PIM risk of patients with more than 15 concurrent medications had a 22.16 times higher PIM risk than those with less than 5 concurrent medications ( OR=22.16, 95% CI: 14.15-34.72, P<0.001). Conclusions:Hospitalized eldery patients with bacterial pneumonia who have more severe comorbidities, longer hospital stay, and multiple concomitant medications are at a higher risk of PIM occurrence. Rational medication use among these patients should be paid attention to in clinical practice.

Objective:To develop and evaluate a deep learning-based automatic measurement model for panoramic ultrasound biomicroscopy (UBM) images.Methods:A diagnostic test study was conducted.Preoperative UBM examination results of 372 patients who underwent implantable collamer lens (ICL) implantation were collected at the Eye Center of Renmin Hospital of Wuhan University between February 2021 and March 2023.A total of 1 368 panoramic UBM images were obtained to establish an image database.The dataset was divided into a training set (760 images), a validation set (86 images) and an internal test set (522 images).An expert panel consisting of three ophthalmologists annotated the images.The UNet+ + network was used to automatically segment anterior segment tissues, such as the cornea, lens and iris.In addition, image processing techniques and geometric localization algorithms were developed to automatically identify the anatomical landmarks of pupil diameter (PD), anterior chamber depth (ACD), angle-to-angle distance (ATA) and sulcus-to-sulcus distance (STS) to complete the measurement of these parameters.Additionally, 480 panoramic UBM images of 135 patients (240 eyes) from Aier Eye Hospital of Wuhan University were used as an external test set to further evaluate the performance of the model in different centers.The consistency between the measurements from the model and expert panel, the Pentacam system was assessed.Finally, 150 images were randomly selected from the external test set for a human-machine comparison to further evaluate the model's performance.This study adhered to the Declaration of Helsinki.The study protocol was approved by the Ethics Committee of Renmin Hospital of Wuhan University (No.WDRY-2022-K109) and Aier eye Hospital of Wuhan University (No.2023IRBKY120903).Written informed consent was obtained from each subject.Results:In the internal test dataset and external test dataset, with manual labeling as the reference standard, the model achieved a mean Dice coefficient of not less than 0.882.At least 95.65% of the anatomical landmark localization results had Euclidean distance differences within 250 μm.The intraclass correlation coefficients (ICCs) for the measurements of PD, ACD, angle-to-angle ATA, and STS were at least 0.958, with mean relative errors not exceeding 2.407%.With the Pentacam measurements as the reference standard, the ICCs for PD in the internal and external test sets were 0.540 and 0.466, respectively, while the ICCs for ACD were 0.946 and 0.908, respectively.In the human-machine comparison, the ICCs between the model's measurements and those of senior experts were all not lower than 0.969.Conclusions:The deep learning-based model can automatically measure anterior segment parameters from preoperative panoramic UBM images of patients undergoing ICL surgery.The model demonstrates a consistency comparable to that of senior experts, while providing higher efficiency.In terms of ACD measurement, the model shows good agreement between the measurements obtained from the model and Pentacam system.

Objective:To evaluate potentially inappropriate medication (PIM) in hospitalized elderly patients with bacterial pneumonia, and explore its influencing factors.Methods:It was a single-center cross-sectional study. The study focused on elderly patients with bacterial pneumonia who were admitted to Xuanwu Hospital, Capital Medical University from January 2018 to November 2022. Patients′ gender, age, weight, length of hospital stay, diagnosis at admission, physical examination, diagnosis at discharge, comorbidities, medications, and laboratory test results were extracted from hospital information system and electronic medical records. Medication use of patients included in the analysis during their hospitalization were evaluated according to the classification of PIMs in the 5 lists of the Beer′s criteria of American Geriatrics Society. Based on whether PIM occurred, the patients were divided into with PIM group and without PIM group. The clinical features between the 2 groups were compared and the influencing factors of PIM were analyzed using multivariable logistic regression.Results:A total of 2 720 patients were included, in which 1 734 (63.75%) were male. The median age was 78 (70, 85) years and their ages ranged from 65 to 103 years. The number of drugs used per patient was 14 (10, 18) kinds, ranging from 1 to 57 kinds. The length of hospital stay was 12 (9, 17) days, ranging from 1 to 162 days. Charlson comorbidity index (CCI) was 6 (5, 8) points. Among the 2 720 patients, 1 894 (69.63%) experienced PIM, with a total of 6 166 cases of PIM. The top 3 drugs ranked by the number of PIM occurrence were antiplatelet agents (1 357 cases), benzodiazapine receptor agonists (956 cases), and antipsychotics (884 cases). The comparison of clinical characteristics between the 2 groups showed that differences in age, CCI, length of hospital stay, and number of medications between with PIM and without PIM patients were statistically significant (all P<0.001). Multivariable logistic regression results showed that CCI, length of hospital stay, and number of medications were independent influencing factors for PIM. The risk increased by 8% and 1% with one point increase in CCI and one day extension in length of hospital stay [odds ratio ( OR)=1.08, 95% confidence interval ( CI): 1.04-1.13, P<0.001; OR=1.01, 95% CI: 1.00-1.03, P=0.03]. PIM risk of patients with more than 15 concurrent medications had a 22.16 times higher PIM risk than those with less than 5 concurrent medications ( OR=22.16, 95% CI: 14.15-34.72, P<0.001). Conclusions:Hospitalized eldery patients with bacterial pneumonia who have more severe comorbidities, longer hospital stay, and multiple concomitant medications are at a higher risk of PIM occurrence. Rational medication use among these patients should be paid attention to in clinical practice.

Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

ObjectiveTo investigate the efficacy and safety of cinobufagin tablets combined with thalidomide/dexamethasone （TD） regimen in the treatment of newly diagnosed multiple myeloma （NDMM） with phlegm and stasis obstruction. MethodsThe clinical data of 50 patients with NDMM of phlegm and stasis obstruction who were hospitalized at the Jiangsu Province Hospital of Chinese Medicine from June 1st， 2015 to July 31th， 2019 were retrospectively analyzed， and they were divided into a control group （bortezomib/dexamethasone-containing regimen， 27 cases） and an observation group （cinobufagin tablets combined with TD regimen， 23 cases）. The clinical efficacy and safety were compared between the two groups after two or three courses of treatment. The primary outcomes were clinical remission rate including overall response rate and deep remission rate， one-year and two-year overall survival rate， and adverse effects. The secondary outcomes were the proportion of plasma cells in bone marrow， hemoglobin， β2-microglobulin， lactate dehydrogenase， serum creatinine， blood urea nitrogen， bone pain score， and KPS functional status score （KPS score） before and after treatment. ResultsIn terms of clinical efficacy， there was no statistically significant difference （P＞0.05） in the overall response rate ［the observation group 69.57%（16/23） vs the control group 70.37% （19/27）］ and deep remission rate ［the observation group 56.52% （13/23） vs the control group 55.56% （15/27）］ between groups after the treatment. The one-year overall survival rates of the observation group and the control group were 90.9% and 92.4%， and the two-year overall survival rates were 81.8% and 80.9% respectively， with no statistically significant differences between groups （P＞0.05）. During the treatment， no renal function injury occurred in both groups. The incidence of peripheral nerve injury in the observation group was 8.70%， which was lower than 48.15% in the control group （P＜0.01）. After the treatment， the proportion of myeloma plasma cells， β2-microglobulin， serum creatinine level， and bone pain score decreased， while the hemoglobin level and KPS score increased in both groups （P＜0.05 or P＜0.01）. Compared between groups after treatment， the bone pain score of the observation group was lower than that of the control group， while the KPS score was higher than that of the control group （P＜0.05）. ConclusionThe clinical efficacy of cinobufagin tablets combined with TD in the treatment of NDMM is equivalent to bortezomib/dexamethasone-containing regimen， but the former is more helpful in relieving the pain and improving the quality of life， and has better safety.

Objective To standardize the management of narrative pharmaceutical documents,and to promote the standardization of narrative pharmaceutical services.Methods With the help of PaaS cloud computing technology design and development,and with the help of OpenAI technology,narrative material writing and narrative communication activities were assisted.Results The functions of narrative prescription,management of narrative activities,narrative writing and parallel medical records were realized and the resource database of narrative materials was established.Conclusion PaaS cloud computing technology has the advantages of rapid deployment and development of narrative pharmaceutical management forms,and it is combined with OpenAI technology.It can help clinical pharmacists quickly complete the search and writing of clinical cases and narrative materials,and greatly improve the efficiency of narrative pharmaceutical services.

Objective To investigate the knowledge,attitude and practice(KAP)of medical staff and parents of children on the risk of using traditional Chinese patent medicine,analyze the similarities and differences between the two groups of people in their knowledge of traditional Chinese patent medicine,medication attitudes and medication behaviors,and analyze the influencing factors of traditional Chinese patent medicine medication risk from the perspective of"doctor-patient",so as to better guide clinical work.Methods From July to November 2023,the medical staff of Children's Hospital Affiliated to the Capital Institute of Pediatrics(hereinafter referred to as"our hospital")and other hospitals,as well as the parents of children who had visited our hospital and taken traditional Chinese patent medicine were taken as the subjects of the survey.The questionnaires were distributed and collected by Questionnarie Stars to analyze the KAP scores of medical staff and parents of children on the risk of children using traditional Chinese patent medicine,and the factors influencing the scores of KAP of taking traditional Chinese patent medicine were analyzed using Logistic regression analysis.And Spearman correlation analysis was used to explore the correlation among knowledge,attitude and practice.Results A total of 339 valid questionnaires(the effective recovery rate of 98.83%)were collected from the medical staff version of the questionnaire.The medical staff with excellent knowledge,attitude and practice scores accounted for 16.22%,7.08%and 83.19%,respectively.A total of 336 valid questionnaires(the effective recovery rate of 98.82%)were collected from the parents'version.The parents of the children with excellent knowledge,attitude and practice scores accounted for 25.87%,3.57%and 30.65%,respectively.Logistic regression analysis showed that different hospital locations and educational levels were important influencing factors for the KAP of medical personnel,while the age,education level,work status,occupation,and monthly income of parents were important influencing factors for their KAP(P<0.05).The Spearman correlation analysis results showed a significant positive correlation among medication knowledge,attitude,and behavior(P<0.01).Conclusion Medical staff and parents of children need to further improve their knowledge and attitude towards the use of traditional Chinese patent medicine.Parents need to pay special attention to standardizing drug use practice,so as to reduce the risk of children using traditional Chinese patent medicine.Physicians and pharmacists can carry out appropriate traditional Chinese patent medicine knowledge popularization and science popularization for parents of children combined with the results of this study.