OBJECTIVE: To develop an artificial intelligence-assisted system for the automatic detection of the features of common 21 auricular points based on the YOLOv8 neural network. METHODS: A total of 660 human auricular images from three research centers were collected from June 2019 to February 2024. The rectangle boxes and features of images were annotated using the LabelMe5.3.1 tool and converted them into a format compatible with the YOLO model. Using these data, transfer learning and fine-tuning training were conducted on different scales of pretrained YOLO neural network models. The model's performance was evaluated on validation and test sets, including the mean average precision (mAP) at various thresholds, recall rate (recall), frames per second (FPS) and confusion matrices. Finally, the model was deployed on a local computer, and the real-time detection of human auricular images was conducted using a camera. RESULTS: Five different versions of the YOLOv8 key-point detection model were developed, including YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x. On the validation set, YOLOv8n showed the best performance in terms of speed (225.736 frames per second) and precision (0.998). On the external test set, YOLOv8n achieved the accuracy of 0.991, the sensitivity of 1.0, and the F1 score of 0.995. The localization performance of auricular point features showed the average accuracy of 0.990, the precision of 0.995, and the recall of 0.997 under 50% intersection ration (mAP⁵⁰). CONCLUSION The key-point detection model of 21 common auricular points based on YOLOv8n exhibits the excellent predictive performance, which is capable of rapidly and automatically locating and classifying auricular points.
Humans ; Neural Networks, Computer ; Artificial Intelligence ; Acupuncture Points

Objective To investigate the clinical significance of neutrophil to albumin ratio(NAR),neutrophil to lymphocyte ratio(NLR),monocyte to lymphocyte ratio(MLR),platelet to lymphocyte ratio(PLR)and systemic immune-inflammation index(SII)in schizophrenia.Methods A total of 215 patients with schizophrenia admitted to Nanning Social Welfare Hospital from January 2019 to June 2021 were included in schizophrenia group,and 113 healthy people who underwent physical examination in hospital during the same period were included in control group.NAR,NLR,MLR,PLR,SII of two groups were compared,and the correlation between them and the clinical symptom score of schizophrenia and changes after taking antipsychotic drugs were analyzed.Results The levels of NAR,NLR,PLR,MLR and SII in schizophrenia group were significantly higher than those in control group(P<0.05).Spearman correlation analysis showed that NAR,NLR,SII and scale for the assessment of positive symptom scores were negatively correlated(P<0.05).After treatment with antipsychotic drugs,all inflammatory indexes were decreased in schizophrenia group,but NAR and MLR were still significantly higher than those in control group(P<0.05).Conclusion There is an active state of immune inflammation in schizophrenia,and the levels of NAR and MLR do not decrease significantly after treatment with antipsychotic drugs,which may be related to the persistence of schizophrenia symptoms.

Objective To investigate the real-world factors influencing the clinical outcome of Febrile neu-tropenia(FN)in oncology patients.Methods We conducted a retrospective analysis of clinical data from 130 FN patients admitted to our hospital between January 2020 and December 2022.The patients were categorized into three groups based on their clinical efficacy:cured group,effective group,and ineffective group.A comparison was made among the three groups regarding general data and laboratory examination results.Univariate and ordered multicategorical logistic regression analyses were performed to identify factors affecting the clinical efficacy of FN.Results The overall effective rate of FN in our hospital was 86.15%.Univariate analysis revealed statistically significant differences among the three patient groups regarding the duration of granulomatous defects,Physical Status Score(PS Score),procalcitonin(PCT)levels,and timing of administration(P＜0.05).Ordinal multicat-egorical logistic regression analysis demonstrated that patients with PS scores＜2,granulomatous defects lasting less than 7 days,and PCT levels below 0.5 ng/mL exhibited better clinical treatment outcomes.Conclusion In the management of FN,it is crucial to prioritize patients with high PS scores and elevated PCT levels while optimizing drug utilization to enhance clinical efficacy.Timely intervention should be implemented to address granulopathy and improve overall clinical outcomes.

Objective Given the extensive application of machine learning(ML)in medical models and its remarkable learning and generalization capabilities,this study employed automated ML(AutoML)combined with patient demographics and clinical conditions to early assess the risk of failure in bowel preparation prior to colonoscopy.Methods A retrospective analysis was conducted on patients who underwent colonoscopy examinations in Hospital 1 and Hospital 2 from January 2022 to January 2023,and their general and clinical information was collected.According to the Boston bowel preparation scale(BBPS),a BBPS of≤5 was defined as a failure in bowel preparation,>5 was deemed satisfactory.From the data of the two hospitals,we randomly divided the dataset into a training set(n=303)and a validation set(n=76)at an 8∶2 ratio.Least absolute shrinkage and selection operator(LASSO)logistic regression(LR)model was used for feature selection,a nomogram scoring system was constructed,and models were established using AutoML based on five algorithms.Model performance was evaluated through receiver operator characteristic curve(ROC curve),calibration curves,LR-based decision curve analysis(DCA),SHAP plots,and force plots.Results Among the 379 patients,105 cases(27.7%)experienced bowel preparation failure(BBPS≤5).21 study variables were narrowed down to 10 through LASSO with 5-fold cross-validation,resulting in the development of a Nomogram chart with demonstrated reliability via calibration curves.Using the H2O platform and five algorithms[gradient boosting machine(GBM),deep learning(DL),generalized linear model(GLM),Stacked Ensemble and distributed random forest(DRF)],67 models were developed.Stacked Ensemble outperformed the others with an area under the curve(AUC)of 0.871,LogLoss of 0.403,and RMSE of 0.354,surpassing traditional LR model and other models.Variable importance contribution plots indicated significant predictive influences from factors such as the interval between laxative ingestion and examination,history of constipation,completion of laxative regimen,age,and presence of a companion during the procedure.Finally,SHAP plots and force plots revealed variable distribution patterns in binary classification predictions and the impact of variables on predictive outcomes.Conclusion The AutoML model based on the Stacked Ensemble algorithm exhibits clear clinical utility in early prediction of bowel preparation failure risk.Moreover,a clinically applicable column chart scoring tool is constructed.

Bacillus subtilis is recognized as a generally-regarded-as-safe strain, and has been widely used in the biosynthesis of high value-added products, including N-acetylneuraminic acid (NeuAc) which is widely used as a nutraceutical and a pharmaceutical intermediate. Biosensors responding to target products are widely used in dynamic regulation and high-throughput screening in metabolic engineering to improve the efficiency of biosynthesis. However, B. subtilis lacks biosensors that can efficiently respond to NeuAc. This study first tested and optimized the transport capacity of NeuAc transporters, and obtained a series of strains with different transport capacities for testing NeuAc-responsive biosensors. Subsequently, the binding site sequence of Bbr_NanR responding to NeuAc was inserted into different sites of the constitutive promoter of B. subtilis, and active hybrid promoters were obtained. Next, by introducing and optimizing the expression of Bbr_NanR in B. subtilis with NeuAc transport capacity, we obtained an NeuAc-responsive biosensor with wide dynamic range and higher activation fold. Among them, P535-N2 can sensitively respond to changes in intracellular NeuAc concentration, with the largest dynamic range (180-20 245) AU/OD. P566-N2 shows a 122-fold of activation, which is 2 times of the reported NeuAc-responsive biosensor in B. subtilis. The NeuAc-responsive biosensor developed in this study can be used to screen enzyme mutants and B. subtilis strains with high NeuAc production efficiency, providing an efficient and sensitive analysis and regulation tool for biosynthesis of NeuAc in B. subtilis.
N-Acetylneuraminic Acid/metabolism* ; Bacillus subtilis/metabolism* ; Promoter Regions, Genetic/genetics* ; Binding Sites ; Biosensing Techniques

Objective    To analyze the correlation between folate receptor-positive circulating tumor cells (FR+CTC) and the benign or malignant lesions of the lung, and to establish a malignant prediction model for pulmonary neoplasm based on clinical data, imaging and FR+CTC tests. Methods    A retrospective analysis was done on 1 277 patients admitted to the Affiliated Hospital of Qingdao University from September 2018 to December 2019, including 518 males and 759 females, with a median age of 57 (29-85) years. They underwent CTC examination of peripheral blood and had pathological results of pulmonary nodules and lung tumors. The patients were randomly divided into a trial group and a validation group. Univariate and multivariate analyses were performed on the data of the two groups. Then the nomogram prediction model was established and verified internally and externally. Receiver operating characteristic (ROC) curve was used to test the differentiation of the model and calibration curve was used to test the consistency of the model. Results    Totally 925 patients suffered non-small cell lung cancer and 113 patients had benign diseases in the trial group; 219 patients suffered non-small cell lung cancer and 20 patients had benign diseases in the verification group. The FR+CTC in the peripheral blood of non-small cell lung cancer patients was higher than that found in the lungs of the patients who were in favorite conditions (P<0.001). Multivariate analysis showed that age≥60 years, female, FR+CTC value>8.7 FU/3 mL, positive pleural indenlation sign, nodule diameter, positive burr sign, consolidation/tumor ratio<1 were independent risk factors for benign and malignant lung tumors with a lesion diameter of ≤4 cm. Thereby, the nomogram prediction model was established. The area under the ROC curve (AUC) of the trial group was 0.918, the sensitivity was 86.36%, and the specificity was 83.19%. The AUC value of the verification group was 0.903, the sensitivity of the model was 79.45%, and the specificity was 90.00%, indicating nomogram model discrimination was efficient. The calibration curve also showed that the nomogram model calibration worked well. Conclusion    FR+CTC in the peripheral blood of non-small cell lung cancer patients is higher than that found in the lungs of the patients who carry benign pulmonary diseases. The diagnostic model of clinical stage Ⅰ non-small cell lung cancer established in this study owns good accuracy and can provide a basis for clinical diagnosis.

Objective:To analyze the characteristics and prognosis of visual field of Leber hereditary optic neuropathy (LHON) with G11778A mutation.Methods:A retrospective clinical study. Twenty-two (44 eyes) of LHON patients diagnosed with G11778A site mutation by mt-DNA examination from May 2008 to February 2018 in Ophthalmology Department of Dongfang Hospital of Beijing University of Chinese Medicine, were enrolled in this study. All patients underwent best corrected visual acuity (BCVA), visual field and optical coherence tomography (OCT). The BCVA examination was performed using the international standard visual acuity chart, which was converted into logarithm of the minimum angle of resolution (logMAR) BCVA for record. The thickness of the retinal nerve fiber layer (RNFL) in the 200μm×200μm annular region 1.73 mm outside the optic disc was measured by OCT. At least 7 visual field examinations were performed within one month before and after 2, 4, 8, 12, 18, 24 and 30 months of the course of disease by using Octopus 101 perimetry. Among 44 eyes, 27 eyes were detected with G2 procedure (G2 group) and 17 eyes were detected with LVC procedure (LVC group). The mean field defect (MD) and mean optical sensitivity (MS) were used as the main outcome indexes. According to the onset age, the patients were further divided into the ≤14 years old group and>14 years old group. There was a significant difference in initial logMAR BCVA between the G2 group and LVC group ( t=4.994, P=0.000), but there was no significant difference in gender ( χ2=1.896, P=0.169) and age ( t=0.337, P=0.708) between the two groups. Independent sample t test was used for comparison between groups, paired t test was used for comparison within groups, and one-way analysis of variance was used for comparison between groups. The statistical data were compared by χ2 test. Results:In the G2 group, the MD value of the subgroup of children (≤14 years old) decreased gradually during the follow-up period, and the MD value since 18 months after onset was significantly lower than the value of 2 months after onset ( t=3.813, 4.590, 5.033; P=0.002, 0.001, 0.000). No obvious visual field index changes were seen in other subgroups ( P>0.05). The central scotoma was the most common type of visual field defect in the early stage, and the diffuse defect was the most common type of visual field defect in the late stage. There was a significant difference in the types of visual field distribution between the early and late stage in G2 group ( χ2=17.414, P=0.015). There was no significant difference in the type of visual field distribution between the early and late stage in LVC group ( χ2=4.541, P=0.474). The MD value in the G2 group remained stable within 8 months after onset, but significantly improved after 18 months after onset ( t=2.100, 3.217, 3.566; P=0.046, 0.003, 0.001). The MS in the LVC group did not significantly improve during follow-up ( P>0.05). The average visual acuity of the G2 group was significantly improved from 12 months ( t=3.039, 3.678, 4.264, 5.078; P=0.008, 0.002, 0.001, 0.000). The visual acuity of the eyes in the G2 group was better than that of the LVC group during all follow-up periods ( P≤0.05). The RNFL thickness of all patients continued to decrease after onset, but the RNFL thickness was significantly higher at 4, 8, 18, 24, 30 months in the G2 group than those in the LVC group ( t=2.471, 2.269, 2.474, 2.509, 2.782; P=0.018, 0.028, 0.017, 0.016, 0.008). Conclusions:The main types of visual field defect of LHON with G11778A mutation are the central scotoma in the early stage, while the diffuse defect and central scotoma are both very common in the later stage. The visual field of LHON patients examined by G2 procedure is significantly improved during the follow-up, as well as the visual acuity improved significantly, and the visual field improvement in younger cases (≤14 years old) is better than that of older cases (>14 years old), but the visual field of the LVC procedure cases did not improve during follow-up.

Objective To analezk thk charactkristics of drug-induckd livkr injure( DIFI)in childrkn with acutk lemphoblastic lkuckmia(LFF),so as to improvk thk phesician＇s undkrstanding of chkmothkrape DIFI,and to guidk clinical rational drug usk. Methods Onk hundrkd and forte-thrkk casks with LFF diagnoskd in thk Dkpartmknt of Hk-matologe and Oncologe in thk Pirst Lffiliatkd Hospital of Yhkngzhou Rnivkrsite from Januare 2012 to Dkckmbkr 2016 wkrk analezkd rktrospkctivkle. Baskd on DIFI diagnostic critkria and thk ARCLM scalk,thk casks with a scork of ≥3 points wkrk considkrkd to havk chkmothkrape DIFI. Groupkd be gkndkr,agk,immunoteping,risc and stagk of chkmo-thkrape,thk incidknck of DIFI was comparkd. Thk situation aftkr DIFI prkvkntion was comparkd bktwkkn two groups which was groupkd according to whkthkr thk application of hkpatoprotkctivk drugs. ResuIts Onk hundrkd and kight ca-sks(75. 52﹪)had DIFI,66 casks(61. 11﹪)showkd clinical manifkstations of livkr injure,and 42 casks(38. 89﹪) had no clinical semptoms. Lmong all thk casks 57. 41﹪(62 casks)wkrk mild livkr damagk,25﹪(27 casks)wkrk modkratk livkr injure and 17. 59﹪(19 casks)wkrk skvkrk livkr damagk. Thk clinical tepks which wkrk hkpatockllular accounting for 79. 63﹪(86 casks),cholkstatic 7. 41﹪(8 casks)and mixkd 12. 96﹪(14 casks). Malk wkrk 80 casks (79. 21﹪)and fkmalk 28 casks(66. 67﹪),but thk incidknck of DIFI bktwkkn diffkrknt gkndkr group had no statistical diffkrknck(χ2 ﹦2. 524,P﹦0. 112). Skvknte-fivk casks(77. 32﹪)wkrk ＜7 ekars agk and 33 casks(71. 74﹪)≥7 ekars agk,and thk incidknck of DIFI bktwkkn 2 groups was not statisticalle diffkrknt(χ2 ﹦0. 526,P﹦0. 468). Thkrk was no significant diffkrknck in T-LFF(8 casks,61. 54﹪)and B-LFF(100 casks,76. 92﹪)( χ2 ﹦0. 795,P﹦0. 372). Thk incidknck had significant diffkrknck in diffkrknt risc(P﹦0. 002). Thk incidknck of DIFI bktwkkn thk middlk risc group(60 casks,88. 24﹪)and standard risc(21 casks,58. 33﹪)had statistical diffkrknck( P ＜0. 05 ). Thk incidknck of DIFI bktwkkn thk middlk risc group and skvkrk risc(27 casks,69. 23﹪)had statistical diffkrknck( P﹦0. 015). Thk incidknck was diffkrknt in diffkrknt stagks of chkmothkrape(P＜0. 05). Thk incidknck of DIFI in induckd stagk was diffkrknt comparkd to othkr stagks(P＜0. 05). ARCLM scork ＞8 points accountkd for 21 casks(19. 45﹪), 6-8 points accountkd for 59 casks(54. 63﹪)and 3 -5 points accountkd for 28 casks(25. 92﹪). Eighte -nink patiknts(92. 71﹪)wkrk kffkctivk in thk hkpatoprotkctivk group and 8 patiknts(66. 67﹪)in thk no hkpatoprotkctivk thkrape group. Thk diffkrknck bktwkkn thk 2 groups was statisticalle significant(χ2 ﹦5. 317,P﹦0. 021). ConcIusions Thk clinical semptoms of drug-induckd livkr injure in childrkn with LFF chkmothkrape ark lacc of spkcificite. Thke ark mainle charactkrizkd be mild livkr injure. Thk clinical tepk of hkpatic injure is common in hkpatockllular. Thk ARCLM scork was mostle 6 to 8. Thkrk is no rklationship bktwkkn thk incidknck in LFF and gkndkr,agk,tepk of lkuck-mia. Thk incidknck with modkratk risc tepk is highkr than that of thk standard and high-risc tepk. Thk incidknck in induction rkmission stagk is highkst. Lpplication of hkpatoprotkctivk drugs is bknkficial to DIFI prognosis.

Objective To investigate the medium- and long-term effects and safty of Pipeline embolization device ( PED) for the treatment of large and giant intracranial anterior circulation aneurysms. Methods From December 2014 to December 2016,the data of 36 consecutive patients with large and giant aneurysm ( 36 large and giant intracranial anterior circulation aneurysms ) treated with PED in Donglei Brain Doctor Group were analyzed retrospectively. The diameter of the aneurysms was 12 -33 mm, (mean16.6±4.5mm),andthenecksizewas4-10mm(mean6.1±1.5mm).Eightaneurysmswere located in the carotid cavernous sinus segment,22 in the ophthalmic artery segment,5 in the internal carotid artery posterior communicating segment, and 1 in the M1 segment of middle cerebral artery. Seven aneurysms were only treated by PED,28 aneurysms were treated by PED in combination with coil embolization,and 1 aneurysm was treated by double PEDs. The modified Rankin scale ( mRS) score was used to evaluate the prognosis of the patients. Results (1) The clinical prognosis of the patients was followed up by telephone and outpatient department for 6-33 months. Twenty-five patients were followed up by DSA,23 aneurysms (92%) were occluded totally (Raymond gradeⅠ) and 2 (8％) were occluded near totally ( Raymond grade Ⅱ) . ( 2 ) Seven patients were treated with PED alone. Four patients were cured totally after 6 months follow-up,1 was occluded subtotally,2 were not cured;6 were cured in the last follow-up (33 months),and the other aneurysm was gradually reduced;17 of 28 patients treated with PED in combination with coils received DSA follow-up. They were followed up for 6-8 months. All the aneurysms were totally occluded ( Raymond grade Ⅰ) . 1 aneurysm was treated by 2 PEDs, DSA revealed micro-aneurysm-like development at 8 months after procedure. The aneurysms were basically occluded after 15-month follow-up. (3) MRI confirmed after operation that 10 patients had asymptomatic scattered spotted ischemic foci,4 had cerebral parenchymal hemorrhage,1 of them died,1 recovered well after treatment (mRS 1),and the other 2 were asymptomatic cerebral hemorrhage. The occupying effect of 24 cases disappeared, 8 had obvious improvement,and 3 did not have any obvious change. Conclusions The occlusion rate of the treatment of large and giant intracranial aneurysms with PED was high. The results of medium-term follow-up showed that the occlusion rate of PED in combination with coils in the treatment of aneurysms was higher than that of PED alone. The long-term follow-up results showed that the occlusion rate of patients treated with PED alone (including one or more) was gradually increased with time. The safety of the surgery needs to be further confirmed by a large sample study.

ObjectiveTo investigate the immunity and seroprevalence of hepatitis A and to identify the risk factors of hepatitis A infection in 0-18 year-old children and adolescents in Shanghai.MethodsSubjects were enrolled by stratifying and clustering random sampling method.Questionnaire interview was applied to investigate the socio-demographic and behavioral factors related to hepatitis A virus (HAV),and information on HAV immunization was abstracted from the immunization registration book of each subject.The enzyme-linked immunosorbent assay (ELISA) was used to qualitatively detect HAV IgM and quantitatively measure total HAV antibody in all subjects.Risk factors associated with HAV among the subjects without HAV vaccination were analyzed.ResultsA total of 2431 subjects were enrolled in the present study with negative HAV IgM antibody and total HAV antibody in 1483 subjects were sero-positive with positivity rate of 61％.Total HAV antibody positivity rates were declined with age increasing and were significantly higher in subjects with HAV vaccination than those without HAV vaccination records.Salad food,eating together without food separation in school and endoscopy inspection were risk factors for HAV infection.ConclusionsHAV vaccination strategies remarkably improve the total HAV antibody seropositive rate in children and adolescents in Shanghai.The risk of HAV infection exists if HAV vaccination is not administrated comprehensively.Therefore,strengthening HAV vaccination and health education are important for children and adolescents to prevent and control of hepatitis A in Shanghai.