Objective To explore the construction of a prediction model for hypertensive nephropathy(HN)based on machine learning.Methods A total of 318 hypertensive patients who visited Taizhou First People's Hospital from April 2023 to March 2024 were included and divided into a training set and a validation set at a ratio of 7:3.Least absolute shrinkage and selection operator(LASSO)algorithm was used to select clinical features from the training set,and 12 clinically significant variables were obtained from 18 clinical variables.Based on the Python 3.10 programming language,the training set was used to train the model.Taking the 12 clinically significant indicators were used as input variables,and whether the occurrence of HN was used as the outcome variable.Three machine learning algorithms,namely logistic regression,support vector machine,and artificial neural network,were used to construct prediction models.The test set was used for internal validation of three models.The performance of the models was compared through accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1.Results Among 12 clinically significant variables screened by the LASSO algorithm,cystatin C and urine protein qualitative were found to be the most predictive.The accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1 values of the Logistic regression,support vector machine,and artificial neural network prediction models constructed by machine learning was 0.94,0.96,0.95,0.87,0.91;0.94,0.97,0.96,0.86,0.91;0.91,0.94,0.93,0.80,0.86,respectively.Conclusion Logistic regression,support vector machine,and artificial neural network based on machine learning all have good predictive effects on the progression of hypertensive patients to HN.Among them,the predictive effects of Logistic regression and support vector machine are similar and better than artificial neural network prediction model.

Objective:To explore the split-face phenomenon in patients with bulbar-involved amyotrophic lateral sclerosis (ALS) through clinical and electrophysiological studies.Methods:A total of 52 clinically definite and clinically probable cases of bulbar-involved ALS, diagnosed according to the World Federation of Neurology El Escorial criteria, were retrospectively collected in the Third Central Hospital of Tianjin from September 2019 to November 2022. And 58 patients with idiopathic facial nerve paralysis with onset time≤7 days who visited the Department of Neurology of the Third Central Hospital of Tianjin during the same period were collected as control group. The firm eye closure (FC) score and cheek bulge (CB) score were used to assess the clinical involvement of facial muscles (dividing into facial muscle involvement group and non-facial muscle involvement group) and the presence of the split-face phenomenon (strong eye closure and weak cheek bulging) in ALS patients. The compound muscle action potential (CMAP) amplitudes of the bilateral orbicularis oculi and orbicularis oris muscles were measured using the Nicolet EDX Viking electromyography/evoked potential system. The CMAP amplitude ratio was calculated. The facial nerve electrophysiological differences were compared between ALS patients with bulbar involvement and patients with idiopathic facial nerve paralysis. The analysis of electrophysiological data across various groups was carried out utilizing the Kruskal-Wallis H test, while pairwise comparisons between groups were executed employing the Bonferroni correction method. Additionally, a stepwise binary Logistic regression analysis was implemented to ascertain the factors associated with facial muscle involvement in patients with bulbar-involved ALS. The receiver operating characteristic (ROC) curve was used to assess the diagnostic accuracy of facial nerve electrophysiological testing in diagnosing ALS in the presence of symptoms of facial muscle involvement. Results:Among the 52 ALS patients with bulbar involvement, there were 20 cases (38.5%) with facial muscle involvements, all of which were bilateral; 16 patients (30.8%) exhibited weakness solely in the ability to puff their cheeks, 1 patient (1.9%) presented with weakness exclusively in closing the eyes, and 3 patients (5.8%) experienced weakness in both closing the eyes and puffing the cheeks. The Revised Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) score of the facial muscle involvement group was lower compared to the non-facial muscle involvement group (36.90±9.20 vs 40.75±5.21, t=2.419, P=0.019), while the FC score and CB score were higher in the facial muscle involvement group [FC score: 0(0, 1) vs 0(0, 0), U=5.854, P<0.001; CB score: 4(3, 4) vs 0(0, 0), U=9.069, P<0.001], showing statistically significant differences. There was no statistically significant difference in the CMAP amplitude of the orbicularis oculi muscle between the facial muscle involvement group and the healthy side of the idiopathic facial nerve paralysis group, the affected side of the idiopathic facial nerve paralysis group, and the non-facial muscle involvement group (all P>0.05). The CMAP amplitude of the orbicularis oris muscle in the facial muscle involvement group [1 100.00 (775.00, 1 375.00) μV] was lower than that in the healthy side of the idiopathic facial nerve paralysis group [1 800.00 (1 400.00, 2 300.00) μV] and the non-facial muscle involvement group [1 555.00 (1 202.50, 1 980.00) μV], with statistically significant differences ( H=5.884, P<0.001; H=4.114, P<0.001). There was no statistically significant difference in the CMAP amplitude of the orbicularis oris muscle between the facial muscle involvement group and the affected side of the idiopathic facial nerve paralysis group ( P>0.05). The CMAP amplitude ratio of the orbicularis oculi/orbicularis oris muscles in the facial muscle involvement group [0.83(0.51, 1.14)] was higher than that in the healthy side of the idiopathic facial nerve paralysis group [0.55(0.39, 0.73)], the affected side of the idiopathic facial nerve paralysis group [0.57(0.40, 0.73)], and the non-facial muscle involvement group [0.60(0.42, 0.71)], with statistically significant differences ( H=-3.440, P=0.003; H=-3.433, P=0.004; H=-3.225, P=0.008). Logistic regression analysis revealed that the CMAP amplitude of orbicularis oris muscle ( OR=0.998,95% CI 0.997-0.999, P<0.001) and ALSFRS-R score ( OR=0.916,95% CI 0.857-0.979, P=0.010) were factors associated with facial muscle involvement in ALS patients with bulbar involvement. The ROC curve analysis results showed that the area under the curve (AUC) of the orbicularis oculi muscle CMAP was 0.629, the AUC of the orbicularis oris muscle CMAP was 0.838, and the AUC of the CMAP amplitude ratio of the orbicularis oculi/orbicularis oris muscles was 0.690 in the facial muscle involvement group. Conclusions:Patients with bulbar-involved ALS have split-face phenomenon characterized by strong eye closure and weak cheek bulging. When bulbar-involved ALS patients have symptoms of facial muscle involvement, the CMAP amplitude of the orbicularis oris muscle decreases significantly, whereas the CMAP amplitude of the orbicularis oculi muscle remains relatively stable, further illustrating the split phenomenon.

Objective:To analyze the clinical and electrophysiological characteristics of patients with sensory neuronopathies (SNN), and to evaluate the significance of electrophysiological markers in the diagnosis and assessment of disease progression.Methods:A retrospective analysis was performed to evaluate the clinical manifestations, electrophysiological characteristics, and spinal cord magnetic resonance imaging (MRI) features of 8 cases diagnosed with SNN at the Third Central Hospital of Tianjin between 2015 and 2023. The neurophysiological examination mainly included limb nerve conduction study (NCS), same core needle electrode electromyography, somatosensory evoked potential (SEP), skin sympathetic reflex (SSR), and contact heat evoked potential (CHEP).Results:Among the 8 cases with SNN, 7 cases exhibited asymmetrical onset and a non-length-dependent pattern. All the 8 cases presented with severe deep sensory ataxia, accompanied by superficial sensory abnormalities and tendon areflexia. Paraneoplastic SNN were the most prevalent etiological subtype (4 cases), all of whom presented peripheral neuropathy as the initial symptom. Among these 4 cases, malignancies were identified in 3 cases and 3 cases presented with anti-Hu antibodies. Among the remaining 4 patients, 2 cases were autoimmune and the other 2 cases were idiopathic. NCS results of the 8 cases revealed decrease or absence of sensory nerve action potential (SNAP) amplitudes, with normal sensory conduction velocities. Six cases showed abnormal SEP, including 2 cases of central damage and 4 cases of peripheral damage, 5 cases had abnormal SSR, and 2 cases exhibited abnormal CHEP. Motor nerve conduction studies were normal in all 8 cases. Six patients underwent spinal MRI, and 4 exhibited abnormal signals in dorsal columns.Conclusions:The hallmark clinical manifestation of SNN is sensory ataxia, characterized by substantial impairment of superficial sensation, typically manifesting in a non-length-dependent distribution. Beyond the widespread and significant reduction in SNAP amplitudes, SNN may also exhibit additional electrophysiological impairments, such as those observed in SEP, SSR and CHEP.SEP combined with spinal cord MRI can improve the detection rate for damages in the central sensory conduction pathway.

Background and Aims:Papillary thyroid carcinoma(PTC),the most common type of thyroid cancer,has been rapidly increasing in incidence worldwide,posing a serious threat to individual health and public healthcare systems.Exposure to low-to-moderate doses of ionizing radiation is more relevant to the daily lives of the general population and,therefore,raises greater public health concerns.It has also been widely recognized as a potential factor in immune system remodeling.This study was conducted to investigate the impact of low-to-moderate dose ionizing radiation on the tumor immune microenvironment of PTC,aiming to reveal the potential hazards of such radiation exposure in PTC patients.Methods:Two datasets(GSE29265 and GSE35570)containing RNA-seq data and corresponding clinical information were retrieved and downloaded from the GEO database.These datasets included thyroid cancer samples from patients exposed to ionizing radiation due to the Chernobyl disaster,as well as sporadic thyroid cancer cases.After data cleaning,merging,batch effect correction,differential gene expression analysis,functional enrichment analysis,immune cell infiltration analysis,and tumor microenvironment analysis were performed using R language.Results:In tumor samples,the radiation-exposed group exhibited significant differential gene expression compared to the sporadic group,with three genes upregulated and 27 genes downregulated.These differentially expressed genes were primarily enriched in biological functions closely related to immune responses,including chemokine activity,immune cell chemotaxis,and tumor immunity.Immune cell infiltration analysis indicated that radiation exposure had a limited impact on immune cell infiltration in normal samples.However,in tumor samples,the immune and ESTIMATE scores were significantly lower in the radiation-exposed group than in the sporadic group.Further analysis revealed that total T cells,CD4+T cells,CD8+T cells,B cells,and cytotoxic lymphocytes exhibited significantly lower infiltration levels in the tumor microenvironment of the radiation-exposed group than the sporadic group.Conclusion:Although low-to-moderate dose ionizing radiation has a relatively minor impact on normal thyroid tissue,it significantly reduces the infiltration of various immune cell subtypes in the PTC tumor microenvironment.This reduction in immune infiltration may have important implications for disease progression.

Background and Aims:Papillary thyroid carcinoma(PTC),the most common type of thyroid cancer,has been rapidly increasing in incidence worldwide,posing a serious threat to individual health and public healthcare systems.Exposure to low-to-moderate doses of ionizing radiation is more relevant to the daily lives of the general population and,therefore,raises greater public health concerns.It has also been widely recognized as a potential factor in immune system remodeling.This study was conducted to investigate the impact of low-to-moderate dose ionizing radiation on the tumor immune microenvironment of PTC,aiming to reveal the potential hazards of such radiation exposure in PTC patients.Methods:Two datasets(GSE29265 and GSE35570)containing RNA-seq data and corresponding clinical information were retrieved and downloaded from the GEO database.These datasets included thyroid cancer samples from patients exposed to ionizing radiation due to the Chernobyl disaster,as well as sporadic thyroid cancer cases.After data cleaning,merging,batch effect correction,differential gene expression analysis,functional enrichment analysis,immune cell infiltration analysis,and tumor microenvironment analysis were performed using R language.Results:In tumor samples,the radiation-exposed group exhibited significant differential gene expression compared to the sporadic group,with three genes upregulated and 27 genes downregulated.These differentially expressed genes were primarily enriched in biological functions closely related to immune responses,including chemokine activity,immune cell chemotaxis,and tumor immunity.Immune cell infiltration analysis indicated that radiation exposure had a limited impact on immune cell infiltration in normal samples.However,in tumor samples,the immune and ESTIMATE scores were significantly lower in the radiation-exposed group than in the sporadic group.Further analysis revealed that total T cells,CD4+T cells,CD8+T cells,B cells,and cytotoxic lymphocytes exhibited significantly lower infiltration levels in the tumor microenvironment of the radiation-exposed group than the sporadic group.Conclusion:Although low-to-moderate dose ionizing radiation has a relatively minor impact on normal thyroid tissue,it significantly reduces the infiltration of various immune cell subtypes in the PTC tumor microenvironment.This reduction in immune infiltration may have important implications for disease progression.

Objective To explore the construction of a prediction model for hypertensive nephropathy(HN)based on machine learning.Methods A total of 318 hypertensive patients who visited Taizhou First People's Hospital from April 2023 to March 2024 were included and divided into a training set and a validation set at a ratio of 7:3.Least absolute shrinkage and selection operator(LASSO)algorithm was used to select clinical features from the training set,and 12 clinically significant variables were obtained from 18 clinical variables.Based on the Python 3.10 programming language,the training set was used to train the model.Taking the 12 clinically significant indicators were used as input variables,and whether the occurrence of HN was used as the outcome variable.Three machine learning algorithms,namely logistic regression,support vector machine,and artificial neural network,were used to construct prediction models.The test set was used for internal validation of three models.The performance of the models was compared through accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1.Results Among 12 clinically significant variables screened by the LASSO algorithm,cystatin C and urine protein qualitative were found to be the most predictive.The accuracy,area under the receiver operating characteristic curve,recall rate,precision,and F1 values of the Logistic regression,support vector machine,and artificial neural network prediction models constructed by machine learning was 0.94,0.96,0.95,0.87,0.91;0.94,0.97,0.96,0.86,0.91;0.91,0.94,0.93,0.80,0.86,respectively.Conclusion Logistic regression,support vector machine,and artificial neural network based on machine learning all have good predictive effects on the progression of hypertensive patients to HN.Among them,the predictive effects of Logistic regression and support vector machine are similar and better than artificial neural network prediction model.

Objective:To explore the split-face phenomenon in patients with bulbar-involved amyotrophic lateral sclerosis (ALS) through clinical and electrophysiological studies.Methods:A total of 52 clinically definite and clinically probable cases of bulbar-involved ALS, diagnosed according to the World Federation of Neurology El Escorial criteria, were retrospectively collected in the Third Central Hospital of Tianjin from September 2019 to November 2022. And 58 patients with idiopathic facial nerve paralysis with onset time≤7 days who visited the Department of Neurology of the Third Central Hospital of Tianjin during the same period were collected as control group. The firm eye closure (FC) score and cheek bulge (CB) score were used to assess the clinical involvement of facial muscles (dividing into facial muscle involvement group and non-facial muscle involvement group) and the presence of the split-face phenomenon (strong eye closure and weak cheek bulging) in ALS patients. The compound muscle action potential (CMAP) amplitudes of the bilateral orbicularis oculi and orbicularis oris muscles were measured using the Nicolet EDX Viking electromyography/evoked potential system. The CMAP amplitude ratio was calculated. The facial nerve electrophysiological differences were compared between ALS patients with bulbar involvement and patients with idiopathic facial nerve paralysis. The analysis of electrophysiological data across various groups was carried out utilizing the Kruskal-Wallis H test, while pairwise comparisons between groups were executed employing the Bonferroni correction method. Additionally, a stepwise binary Logistic regression analysis was implemented to ascertain the factors associated with facial muscle involvement in patients with bulbar-involved ALS. The receiver operating characteristic (ROC) curve was used to assess the diagnostic accuracy of facial nerve electrophysiological testing in diagnosing ALS in the presence of symptoms of facial muscle involvement. Results:Among the 52 ALS patients with bulbar involvement, there were 20 cases (38.5%) with facial muscle involvements, all of which were bilateral; 16 patients (30.8%) exhibited weakness solely in the ability to puff their cheeks, 1 patient (1.9%) presented with weakness exclusively in closing the eyes, and 3 patients (5.8%) experienced weakness in both closing the eyes and puffing the cheeks. The Revised Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) score of the facial muscle involvement group was lower compared to the non-facial muscle involvement group (36.90±9.20 vs 40.75±5.21, t=2.419, P=0.019), while the FC score and CB score were higher in the facial muscle involvement group [FC score: 0(0, 1) vs 0(0, 0), U=5.854, P<0.001; CB score: 4(3, 4) vs 0(0, 0), U=9.069, P<0.001], showing statistically significant differences. There was no statistically significant difference in the CMAP amplitude of the orbicularis oculi muscle between the facial muscle involvement group and the healthy side of the idiopathic facial nerve paralysis group, the affected side of the idiopathic facial nerve paralysis group, and the non-facial muscle involvement group (all P>0.05). The CMAP amplitude of the orbicularis oris muscle in the facial muscle involvement group [1 100.00 (775.00, 1 375.00) μV] was lower than that in the healthy side of the idiopathic facial nerve paralysis group [1 800.00 (1 400.00, 2 300.00) μV] and the non-facial muscle involvement group [1 555.00 (1 202.50, 1 980.00) μV], with statistically significant differences ( H=5.884, P<0.001; H=4.114, P<0.001). There was no statistically significant difference in the CMAP amplitude of the orbicularis oris muscle between the facial muscle involvement group and the affected side of the idiopathic facial nerve paralysis group ( P>0.05). The CMAP amplitude ratio of the orbicularis oculi/orbicularis oris muscles in the facial muscle involvement group [0.83(0.51, 1.14)] was higher than that in the healthy side of the idiopathic facial nerve paralysis group [0.55(0.39, 0.73)], the affected side of the idiopathic facial nerve paralysis group [0.57(0.40, 0.73)], and the non-facial muscle involvement group [0.60(0.42, 0.71)], with statistically significant differences ( H=-3.440, P=0.003; H=-3.433, P=0.004; H=-3.225, P=0.008). Logistic regression analysis revealed that the CMAP amplitude of orbicularis oris muscle ( OR=0.998,95% CI 0.997-0.999, P<0.001) and ALSFRS-R score ( OR=0.916,95% CI 0.857-0.979, P=0.010) were factors associated with facial muscle involvement in ALS patients with bulbar involvement. The ROC curve analysis results showed that the area under the curve (AUC) of the orbicularis oculi muscle CMAP was 0.629, the AUC of the orbicularis oris muscle CMAP was 0.838, and the AUC of the CMAP amplitude ratio of the orbicularis oculi/orbicularis oris muscles was 0.690 in the facial muscle involvement group. Conclusions:Patients with bulbar-involved ALS have split-face phenomenon characterized by strong eye closure and weak cheek bulging. When bulbar-involved ALS patients have symptoms of facial muscle involvement, the CMAP amplitude of the orbicularis oris muscle decreases significantly, whereas the CMAP amplitude of the orbicularis oculi muscle remains relatively stable, further illustrating the split phenomenon.

Objective:To analyze the clinical and electrophysiological characteristics of patients with sensory neuronopathies (SNN), and to evaluate the significance of electrophysiological markers in the diagnosis and assessment of disease progression.Methods:A retrospective analysis was performed to evaluate the clinical manifestations, electrophysiological characteristics, and spinal cord magnetic resonance imaging (MRI) features of 8 cases diagnosed with SNN at the Third Central Hospital of Tianjin between 2015 and 2023. The neurophysiological examination mainly included limb nerve conduction study (NCS), same core needle electrode electromyography, somatosensory evoked potential (SEP), skin sympathetic reflex (SSR), and contact heat evoked potential (CHEP).Results:Among the 8 cases with SNN, 7 cases exhibited asymmetrical onset and a non-length-dependent pattern. All the 8 cases presented with severe deep sensory ataxia, accompanied by superficial sensory abnormalities and tendon areflexia. Paraneoplastic SNN were the most prevalent etiological subtype (4 cases), all of whom presented peripheral neuropathy as the initial symptom. Among these 4 cases, malignancies were identified in 3 cases and 3 cases presented with anti-Hu antibodies. Among the remaining 4 patients, 2 cases were autoimmune and the other 2 cases were idiopathic. NCS results of the 8 cases revealed decrease or absence of sensory nerve action potential (SNAP) amplitudes, with normal sensory conduction velocities. Six cases showed abnormal SEP, including 2 cases of central damage and 4 cases of peripheral damage, 5 cases had abnormal SSR, and 2 cases exhibited abnormal CHEP. Motor nerve conduction studies were normal in all 8 cases. Six patients underwent spinal MRI, and 4 exhibited abnormal signals in dorsal columns.Conclusions:The hallmark clinical manifestation of SNN is sensory ataxia, characterized by substantial impairment of superficial sensation, typically manifesting in a non-length-dependent distribution. Beyond the widespread and significant reduction in SNAP amplitudes, SNN may also exhibit additional electrophysiological impairments, such as those observed in SEP, SSR and CHEP.SEP combined with spinal cord MRI can improve the detection rate for damages in the central sensory conduction pathway.

Objective To explore the value of dynamic contrast enhanced MRI(DCE-MRI)for preoperative predicting microvascular invasion(MVI)in hepatocellular carcinoma(HCC).Methods Sixty-eight HCC lesions in 66 patients(64 with single and 2 with 2 lesions)confirmed by surgery and pathology were retrospectively enrolled and were divided into MVI-positive group(n=17)and MVI-negative group(n=51)according to pathological results.Preoperative DCE-MRI qualitative and quantitative parameters of lesions were compared between groups,and the independent risk factors related to preoperative DCE-MRI features of HCC MVI were screened by multivariate binary logistic regression analysis,then a combined model was established.Receiver operating characteristic(ROC)curve was plotted,and the area under the curve(AUC)was calculated to evaluate the predictive efficacy of single independent risk factor and combined model.Results DCE-MRI showed that the proportion of irregular shape,incomplete pseudocapsule,intratumoral necrosis,peritumoral satellite nodules,transient hepatic parenchymal enhancement in arterial phase and mosaic appearance in MVI-positive group were all higher than those in MVI-negative group,the maximum diameter of lesions in MVI-positive group was larger than that in MVI-negative group,while portal-phase enhancement washout(PEW),absolute enhancement washout(AEW)and relative enhancement washout(REW)in MVI-positive group were all lower than those in MVI-negative group(all P＜0.05).Peritumoral satellite nodules(OR=33.777,P=0.003),the maximum diameter of lesion≥4.25 cm(OR=6.429,P=0.038)and REW≤0.15(OR=6.148,P=0.028)were all independent risk factors of MVI in HCC.The AUC of the above single independent risk factors and combined model was 0.755,0.719,0.781 and 0.897,respectively.Conclusion DCE-MRI could effectively preoperatively predict MVI in HCC.

Objective:To evaluate the effect of virtual reality arthroscopic simulation system combined with case-based learning (CBL) in orthopedic clinical teaching.Methods:From January 2021 to June 2022, a total of 36 residents who took the standardized residency training at Department of Articular Orthopedics in The Third Affiliated Hospital of Soochow University were enrolled in this study. They were randomized into an experimental group and a control group, with 18 trainees in each group. The experimental group adopted the virtual reality arthroscopic simulation system combined with CBL teaching mode, while the control group adopted simple endoscopic simulation training box with traditional teaching mode. Theoretical and practical tests and questionnaire survey were carried out to evaluate the teaching effects in two groups. SPSS 22.0 was performed to conduct t-test on the measurement data between the two groups. Results:The theoretical and practical test scores of the experimental group [(89.39±3.09) and (82.72±4.28)] were better than those of the control group [(86.22±4.43) and (76.61±5.65)], with statistically significant differences ( t= 2.49, P=0.018; t=3.66, P=0.001). The questionnaire survey showed that the experimental group was better than the control group in creating novel forms of courses [(4.94±0.24) vs. (4.11±0.58), t=5.62, P<0.001], simulating more real scenes [(4.83±0.38) vs. (3.78±0.43), t=7.80, P<0.001], inspiring learning interests [(4.78±0.43) vs. (4.00±0.59), t=4.51, P<0.001], improving practical skills [(4.83±0.38) vs. (3.83±0.51), t=6.61, P<0.001], and building career confidence [(4.50±0.62) vs. (3.06±0.54), t=7.47, P<0.001], with statistically significant differences (all P<0.001). Conclusions:The virtual reality arthroscopic simulation system combined with CBL teaching mode can better simulate the real clinical scenes, help inspire the interest of learning and quickly improve practical skills, thereby improving the effect of orthopedic clinical teaching.