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 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.

Humans ; Proto-Oncogene Proteins p21(ras)/genetics* ; Prognosis ; Biomarkers, Tumor/genetics* ; Mutation/genetics* ; Colorectal Neoplasms/genetics* ; Proto-Oncogene Proteins B-raf/metabolism*

Anti-contactin-1(CNTN1) IgG4 antibody is a reliable biomarker of a specific subset of chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) patients with distinct clinical features, including advanced age, aggressive symptom, predominantly motor involvement, sensory ataxia, and poor response to intravenous immunoglobulin. Nerve conduction study showed decrease of nerve conduction velocity, reduction of compound motor active potential amplitude without temporal dispersion at early stage of the disease. The article reported an anti-CNTN1 IgG4 antibody positive case, reviewed the current relevant literature, and then summarized the clinical characteristics of anti-CNTN1 IgG4 antibody associated CIDP.

Objective: To evaluate the severity of median nerve damage in patients with carpal tunnel syndrome (CTS) , and to analyze its relationship with body mass index (BMI) and wrist joint index. Methods: From May 2016 to January 2017, 23 patients with mild CTS (mild group) and 35 patients with moderate to severe CTS (moderate to severe group) were enrolled in this study. And 22 healthy volunteers matched for sex and age were selected as control group. The neuroelectrophysiological monitor was used to measure the median nerve movement and sensory nerve conduction in the subjects. The BMI and wrist joint index were calculated. The relationship of neuroelectrophysiological parameters with BMI and wrist joint index was analyzed in the CTS patients. Results: Compared with the control group, the mild group and the moderate to severe group had significantly higher wrist joint index, significantly longer distal motor latency (DML) of the median nerve, and significantly lower sensory nerve conduction velocity (SNCV) and sensory nerve action potential (SNAP) amplitude of the finger 1-wrist and finger 3-wrist (P<0.01) ; the moderate to severe group had significantly higher BMI and significantly lower composite muscle action potential (CMAP) amplitude (P<0.01) . The wrist joint index and BMI were positively correlated with DML of the median nerve and negatively correlated with SCV and SNAP amplitude of the finger 1-wrist and finger 3-wrist (all P<0.05) . The patients with a wrist joint index of >0.73 had a significantly higher risk of CTS than those with a wrist joint index of <0.73 (odd ratio=30.67, 95% confidence interval: 3.79-248.36) . Conclusion A wrist joint index of >0.73 is an independent risk factor for CTS in manual laborers. CTS should be prevented in the manual laborers with high wrist joint index and BMI.

Objective To study the electrophysiological characteristics of hands muscle of upper limb onset amyotrophic lateral sclerosis (UL-ALS), and the variant-flail arm syndrome (FAS) for diagnosis and differential diagnosis. Methods We chose 55 UL-ALS and 12 FAS patients as the cases, 20 cervical spondylotic amyotrophy (CSA) patients as the case controls, and 20 healthy volunteers as the normal controls from January 2013 to March 2018 in the Third Central Hospital of Tianjin. Conventional nerve conduction studies of the median nerve and ulnar nerve were performed in all the patients. The main analysis was done in the compound muscle action potential (CMAP) recorded on the abductor pollicis brevis (APB) and abductor digiti minimi (ADM) and the ratio of the two. Results The ratio of CMAPAPB/CMAPADM of ALS was 0.59 (0.25, 0.79), which was depressed obviously compared with FAS (1.02 (0.92, 1.18), Z=-4.440, P=0.000), CSA (1.88 (1.42, 3.19), Z=-5.902,P=0.000) and the normal controls (0.96 (0.88, 1.15), Z=-5.416, P=0.000). The low ratio of CMAPAPB/CMAPADM (<0.6) was encountered in 40％(23/55) ALS patients, 0 CSA patient and 1/12 FAS patients. An absent APB CMAP and an abnormally low APB/ADM CMAP amplitude ratio (<0.25) were observed only in 25.4％ (14/55) ALS patients. The area under receiver operating characteristic curve in patients of UL-ALS was 0.911 (P=0.000), and in FAS was 0.518 (P=0.559). Using a cut-off value of CMAPAPB/CMAPADM=0.7 for diagnosing ALS yielded 85.5％sensitivity and 95.0％specificity. Conclusion The split hand syndrome is not specific for ALS; however, the low APB/ADM CMAP amplitude ratio may help predict prognosis and can be the diagnostic marker for ALS.

This study retrospectively reviewed 75 patients with impaired glucose tolerance(IGT)admitted in our hospital from March 2015 to October 2015.All patients underwent Toronto clinical scoring system(TCSS) evaluation.Patients with IGT were further divided into normal score group(TCSS-N, n=50)and abnormal score group(TCSS-A, n=25)according to their scoring results, and 30 healthy volunteers were served as control group.All patients and controls underwent motor and sensory nerve conduction studies, as well as sympathetic skin response(SSR)test using the Keypoint.Net(Medoc Ltd)electromyogram device.The results showed that the SSR amplitude of lower limbs was reduced [(0.61±0.44 vs 1.00±0.33)mv, P<0.05]andlatencyoflowerlimbswas extended [(1 880±282 vs 1 642±256)ms, P<0.05]in IGT group compared with control group.But median, ulnar, tibial, and peroneal nerve sensory and motor conduction revealed no difference between two groups.In TCSS-A group, the SSR amplitude of lower limbs was reduced [(0.47±0.39)mv, P<0.05], latency of lower limbs was extended [(2 062±291)ms,P<0.05]and the sensory nerve action potential(SNAP)amplitude of the tibial nerve was significantly lower compared with control group [(1.83±0.37 vs 2.07±0.30)μv, P<0.05].Compared to TCSS-N group, latency of lower limbs was extended [(2 062±291 vs 1 808±246)ms, P<0.05]in TCSS-A group.The SSR amplitude of lower limbs were reduced[(0.66±0.44)mv,P<0.05]and latency were prolonged(P<0.05)in TCSS-N group compared with control group.Pearson correlation analysis showed that the SSR amplitude and latency of the lower limbs were correlated with the postprandial blood glucose, blood glucose fluctuation, body weight, as well as body mass index.These results suggest that there exists peripheral nerve damage in the patients with IGT, mainly involving the small fiber nerve of the lower limbs.Large fibers may also be mildly affected as the disease progresses.

Objective To use nerve conduction study (NCS) to evaluate the function of large fibers,skin sympathetic response (SSR) and contact heat evoked potential (CHEP) to evaluate the function of small fibers in patients with impaired glucose regulation (IGR),and to analyze the occurrence of peripheral neuropathy and damage characteristics.Methods According to the classification criteria of glucose metabolism proposed by WHO in 2006,we selected 120 patients with IGR from January 2015 to December 2016 in our hospital,including 37 impaired fasting glucose (IFG) patients,83 impaired glucose tolerance (IGT) patients,and 60 normal subjects served as control group.All subjects received median,ulnar,tibial,peroneal,sural NCS,SSR and CHEP using the Keypoint.net (Medoc Ltd) electromyogram device.IGR patients were evaluated using the Michigan Neuropathy Screening Instrument (MNSI).Results The abnormal rate of MNSI score in IGR patients was 18.3％ (22/120);the abnormal rate of NCS was 22.5％ (27/120),and the abnormal rate of SSR was 39.2％ (47/120).In IFG group,the abovementioned abnormal rates were 8.1％ (3/37),13.5％ (5/37),29.7％ (11/37) respectively,and 22.9％ (19/83),26.5％ (22/83),43.4％ (36/83) in IGT group.Compared with control group,the tibial,sural nerve sensory nerve action potential amplitude decreased in IGT group (1.3 (0.1,1.9) μV vs 1.4(1.1,3.2) μV,Z=-3.05,P=0.002;(10.5±2.0)μV vs (7.6 ± 1.9)μV,t=0.60,P=0.001);and there was no significant difference between IFG patients and control group.Compared with control group,IFG patients' SSR amplitude reduced in lower limbs (0.7 (0.4,0.8) mV vs 0.8 (0.6,1.0) mV,Z =-2.95,P =0.003),CHEP amplitude decreased in dorsum hand and peroneal area stimuli ((63.0 ±10.0)μV vs(52.4 ±15.3)μV,t=0.61,P=0.003;(44.7 ±12.5)μV vs (28.2 ± 10.6)μV,t=0.31,P =0.000);and in IGR group,SSR amplitude reduced in upper and lower limbs (1.1 (0.5,2.2) mV vs 1.3(0.7,2.6)mV,Z=-2.12,P=0.030;0.4(0.2,0.8)mV vs 0.8(0.6,1.0) mV,Z=-5.96,P=0.000),CHEP amplitude decreased in dorsum hand and peroneal area stimuli ((63.0 ± 10.0) μV vs (38.7± 13.5)μV,t =0.37,P=0.000;(44.7 ±12.5)μV vs(21.9 ± 13.6)μV,t =0.35,P=0.000).Conclusions There is peripheral neuropathy in IGR patients,and the incidence of neuropathy in patients with IGT is higher than those with IFG.Neurophysiological methods are earlier than clinical scores to detect neuropathy.There are only small fiber damages in IFG patients,and IGT patients present by large and small fibrous lesions,mainly in small fibers and lower sensory nerve fibers,characterized by axonal damage and length dependence.

Objective To evaluate the impairment of large and small fibers of peripheral nerves in patients with diabetes mellitus according to the current perception threshold ( CPT ) comparing with nerve conduction studies ( NCS) and sympathetic skin response ( SSR).Methods Fourty-nine diabetic patients were enrolled from August to December 2014 in Tianjin Third Central Hospital.The CPT was examined by the Neurometer?CPT/C in diabetic patients and 23 healthy subjects.Electrical stimulations of 2 000, 250 and 5 Hz were delivered respectively to the first toe distal section , the bottom of the knee-patella middle line and the index finger distal section.NCS and SSR were performed in diabetic patients.The patients were divided into normal NCS ( NCSN) group and abnormal NCS ( NCSA) group according to the results of the nerve conduction.According to the results of SSR , the patients were divided into normal SSR ( SSRN) group and abnormal SSR ( SSRA ) group.Results In three testing areas , the CPT in NCSN group had no significant difference compared with control group , which was tested by different frequency;NCSA group had increased CPT (393.5 ±68.4, 213.1 ±39.7, 172.5 ±52.8, 221.6 ±48.0, 91.7 ±33.1, 76.2 ±27.8, 274.0 ±48.4, 120.1 ±26.6, 75.3 ±21.9) compared with control group (278.1 ±53.6, 128.9 ±46.7, 78.7 ±28.7, 149.3 ±54.6, 52.3 ±23.2, 41.9 ±18.7, 214.2 ±46.0, 96.0 ±29.3, 48.6 ±17.0; F=36.277, 40.827, 50.301, 10.240, 12.837, 13.832, 11.749, 5.559, 17.412,all P<0.01).At the first toe section, CPT decreased in SSRN group ( 91.0 ±29.0, 55.2 ±21.3 ) compared with control group (128.9 ±46.7, 78.7 ±28.7;F=23.071, 22.646, both P<0.05), which was tested by the currents of 250 Hz and 5 Hz frequency.In three testing areas , the SSRA group had increased CPT compared with control group , which was tested by different frequency.Conclusion Hyperesthesia as an early manifestation of small fiber damage can be detected by CPT in diabetic patients; CPT has no advantage in the evaluation of large fiber damage.

Objective To study the value of methionine loading test (MLT) in the mild vascular cognitive impairment (VCI) after acute cerebral infarction.Methods The fasting plasma homocystine (Hcy) level and homocystine level after MLT were measured by high-performance liquid chromatography methods.We chose 240 patients with normal level of fasting plasma Hcy (normal group),159 patients with normal level of Hcy after MLT,81 patients with hyperhomocysteinemia after MLT (hyperhomocysteinemia group),and 112 patients with fasting hyperhomocysteinemia (fasting hyperhomocysteinemia group) in this study.The Montreal Cognitive Assessment (MoCA) and Mini-Mental State Examination (MMSE) were conducted in normal,hyperhomocysteinemia and fasting hyperhomocysteinemia groups on admission,at 7d,14 d,30 d after treatment.Results Logistic regression analysis showed that the increased level of Hcy might be an independent risk factor for VCI [OR:1.285,95％CI:1.038-1.265,P＜0.05].The scores of MMSE and MoCA were lower in patients with fasting hyperhomocysteinemia and patients with hyperhomocysteinemia after MLT than in patients with normal fasting plasma Hcy at 7 d,14 d and 30 d after treatment (P＜0.01 or 0.05),while the scores had no significant differences among the three group on admission (P＞0.05).There were no significant differences in MMSE and MoCA scores between patients with fasting hyperhomocysteinemia and patients with hyperhomocysteinemia after MLT on admission,7 d,14 d and 30 d after treatment (P＞0.05).Conclusions Hcy may be an independent risk factor for VCI.The MLT can discover the dormant vascular risk factors for VCI,which offers a valuable detection method for early intervention and prevention in the clinical medicine.