Comparison and development of machine learning tools in the prediction of chronic kidney disease progression.

BACKGROUND: Urinary protein quantification is critical for assessing the severity of chronic kidney disease (CKD). However, the current procedure for determining the severity of CKD is completed through evaluating 24-h urinary protein, which is inconvenient during follow-up. OBJECTIVE: To quickly predict the severity of CKD using more easily available demographic and blood biochemical features during follow-up, we developed and compared several predictive models using statistical, machine learning and neural network approaches. METHODS: The clinical and blood biochemical results from 551 patients with proteinuria were collected. Thirteen blood-derived tests and 5 demographic features were used as non-urinary clinical variables to predict the 24-h urinary protein outcome response. Nine predictive models were established and compared, including logistic regression, Elastic Net, lasso regression, ridge regression, support vector machine, random forest, XGBoost, neural network and k-nearest neighbor. The AU-ROC, sensitivity (recall), specificity, accuracy, log-loss and precision of each of the models were evaluated. The effect sizes of each variable were analysed and ranked. RESULTS: The linear models including Elastic Net, lasso regression, ridge regression and logistic regression showed the highest overall predictive power, with an average AUC and a precision above 0.87 and 0.8, respectively. Logistic regression ranked first, reaching an AUC of 0.873, with a sensitivity and specificity of 0.83 and 0.82, respectively. The model with the highest sensitivity was Elastic Net (0.85), while XGBoost showed the highest specificity (0.83). In the effect size analyses, we identified that ALB, Scr, TG, LDL and EGFR had important impacts on the predictability of the models, while other predictors such as CRP, HDL and SNA were less important. CONCLUSIONS: Blood-derived tests could be applied as non-urinary predictors during outpatient follow-up. Features in routine blood tests, including ALB, Scr, TG, LDL and EGFR levels, showed predictive ability for CKD severity. The developed online tool can facilitate the prediction of proteinuria progress during follow-up in clinical practice.

[The design of the software system for the interpretation and guidance of drug locus related to drug treatment].

Based on the pharmacogenomics theory, this study developed a software system for interpretation of drug gene loci and guidance on clinical safe medication with the purpose of providing clinical guidance on the safety and effectiveness of drug use through accurate and efficient detection and interpretation of drug gene loci. The system infrastructure was built on a service-oriented architecture (SOA) design and Docker container virtualization approach to achieve a rapid and automatic interpretation of genetic results and best available drugs. The front end was established on HTML5 and JavaScript to realize visualization of analysis results and user interaction. The system was tested and validated to show robust performance which is reliable in clinical use. It will show high impact on the development of pharmacogenomics and clinical practice of patients with personalized medicine.

[Implementation of control system and software design for limbs rehabilitation training based on PCI-1240].

This article presents the design of a motion control system for seated lower-limb rehabilitation training. The system is composed of lower limb exoskeleton, motor drive circuit, program of motion control, and so forth. The power of lower limbs joints is provided by six motors. The PCI-1240 motion control card is used as the core. This study achieved repetitive rotation training and gait trajectory training of lower limbs joints, of which the velocity, angle and time can be accurately controlled and adjusted. The experimental results showed that the motion control system can meet the requirement of repetitive rehabilitation training for patients with lower limb dysfunction. This article provides a new method to the research of motion control system in rehabilitation training, which can promote industrial automation technique to be used for health care, and conducive to the further study of the rehabilitation robot.

[The design of weight-loss walking rehabilitation training system of differential-pressure style].

This research is to develop a weight-loss walking rehabilitation training system based on differential air pressure. The system adopted Proportion-Integral-Derivative (PID) algorithm to improve the precision of weight loss, taking MSP430F149 microprocessor of Texas Instruments as the core of pressure control system. The training software is designed based on Microsoft Visual C++ 6.0 of Microsoft. The system can provide comfortable training environment for patients with lower limb motor function impediment, and can collect electromyographic signals from patients, so as to further the scientific and normative management of the patient's information. Based on this training system, the initial bearing weight, bearing weight after maximum weight loss, and maximum weight loss percentage of 10 normal adults' lower limbs were collected. It was found that the intraclass correlation coefficient (ICC) values were all greater than 0.6. The training system has a good reliability, which can provide scientific data for clinical weight-loss lower limb rehabilitation training.

[Study of New Micropore RF system on Lesion Formation and Complications].

OBJECTIVES: To study the safety and effectiveness of a new type of micropore ablation catheter in vitro ablation system, and to provide reference for clinical practice. METHODS: To evaluate two kinds of catheter in cardiac tissue ablation depth, tissue temperature and thrombosis situation by the same RF system. RESULTS: The power set 25 W, There was no significant difference in ablation depth between the two groups, and no Pop and thrombosis occurred. When the power is more than 40 W, two groups occurred more Pop and thrombosis. CONCLUSIONS: When using high power for Cardiac RF ablation, doctors should pay more attention to complications and thrombosis.

[Analysis of the Muscle Fatigue Based on Band Spectrum Entropy of Multi-channel Surface Electromyography].

Exercise-induced muscle fatigue is a phenomenon that the maximum voluntary contraction force or power output of muscle is temporarily reduced due to muscular movement.If the fatigue is not treated properly,it will bring about a severe injury to the human body.With multi-channel collection of lower limb surface electromyography signals,this article analyzes the muscle fatigue by adoption of band spectrum entropy method which combined electromyographic signal spectral analysis and nonlinear dynamics.The experimental result indicated that with the increase of muscle fatigue,muscle signal spectrum began to move to low frequency,the energy concentrated,the system complexity came down,and the band spectrum entropy which reflected the complexity was also reduced.By monitoring the entropy,we can measure the degree of muscle fatigue,and provide an indicator to judge fatigue degree for the sports training and clinical rehabilitation training.

[Development of Muscle Strength Evaluating System Based on Mobile Platform].

The development of muscle strength evaluating system based on Android system was developed in this research. The system consists of a lower unit and an intelligent mobile terminal. The pressure sensor of the lower unit was used to collect muscle strength parameters. And the parameters were sent to the Android device through the wireless Bluetooth serial port. Then the Android device would send the parameters to the doctor monitored platform through the Internet. The system realized analyzing the muscle strength parameters and real-time displaying them. After it ran on the Android mobile phones, it showed an effective result which proved that the system combined with mobile platform could make more convenient for the patients to assess their own muscle strength. It also provided reliable data references for doctors to know the patients' rehabilitation condition and to make the next rehabilitation plan.

[Control System Design and Test of Digital Comprehensive Training Treadmill for Rats].

This article presents the design of a treadmill of comprehensive training experiment for rats. The treadmill is composed of six tracks and two of them were designed as conventional plane, two were designed to swing right and left, and two were designed to swing back and forth. The power was provided by six motors. The MSP430F149 is used as core to adjust the swing rate and the grade of electric shock. The IAR for MSP430 is used to design the software. The speed of the six tracks could be adjusted between 0 and 30 m/min. The swing tracks of back and forth can be swung for 3-25 times per minute and the swing tracks of right and left for 3-32 times. The electric shock can be divided into three levels, i. e. strong, middle, and weak level for each track. The digital comprehensive training treadmill can meet different training needs, and provide experimental data for mechanism research of some related diseases.

[Design of Balance Function Telerehabilitation System Based on C/S].

This article shows a new design of telerehabilitation system for balance function assessment and training in our laboratory. The system is based on C/S network architecture, and realizes the telecommunication through socket network communication technology. It implements the teletransmission of training data and assessment report of sit-down and stand-up, online communication between doctors and patients, and doctors'management of patient information. This system realizes remote evaluation and telerehabilitation of patients, and brings great convenience for the patients.

[Dynamic modeling and simulating of squatting-standing action of human body].

It is very difficult for stroke patients to complete the action of squatting-standing because their equilibrium function ability has been seriously declined. It was necessary, therefore, to do a deep research on the action of human squatting-standing and to set up an accurate model and simulation. In our modeling research, the movements of upper limbs and head was neglected, and a seven-segment model was developed to establish the coordinate system of human squatting-standing action. It calculated the knee joint moment and hip joint moment during squatting and standing by utilizing Lagrange method, and then simulated this mathematical model by utilizing Matlab. Geometric model of human squatting-standing was developed and simulated in ADAMS which proved that the established Lagrange model was reasonable. It would also provide significant theoretical references for further study and development of squatting-standing rehabilitation training equipment.

[Research on measuring the velocity and displacement of the coxa and knee based on video image processing].

Based on repeated experiments as well as continuous researching and improving, an efficient scheme to measure velocity and displacement of the coxa and knee movements based on video image processing technique is presented in this paper. The scheme performed precise and real-time quantitative measurements of 2D velocity or displacement of the coxa and knee using a video camera mounted on one side of the healing and training beds. The beds were based on simplified pinhole projection model. In addition, we used a special-designed auxiliary calibration target, composed by 24 circle points uniformly located on two concentric circles and two straight rods which can rotate freely along the concentric center within the vertical plane, to do the measurements. Experiments carried out in our laboratory showed that the proposed scheme could basically satisfy the requirements about precision and processing speed of such kind of system, and would be very suitable to be applied to smart evaluation/training and healing system for muscles/balance function disability as an advanced and intuitional helping method.

Demonstration of posturographic parameters of squat-stand activity in hemiparetic patients on a new multi-utility balance assessing and training system.

BACKGROUND: Quantitative evaluation of position control ability in stroke patients is needed. Here we report a demonstration of position control ability assessment and test-retest reliability during squat-stand activity on a new system in hemiparetic patients and controls. METHODS: Sixty-two healthy adults and thirty-four hemiparetics were enrolled. RESULTS: During partial weight support, the ICCs ranged from 0.77 to 0.91, which indicated a good reliability. During standard weight bearing and resistance, the ICCs varied from 0.64 to 0.86 and 0.54 to 0.84, respectively, indicating a fair reliability. Compared with the healthy adults, the stroke patients demonstrated poorer position control ability. CONCLUSIONS: The posturography of the squat-stand activity is a new and reliable measurement tool for position control. According to the methods proposed here, hemiparetics can be differentiated from healthy adults using the squat-stand activity. This activity will provide a new evaluation tool and therapy with visual feedback for the stroke patients. TRIAL REGISTRATION: Chinese clinical trial registry, ChiCTR-TRC-10000863.

[Development of a testing device for knee joint kinematics parameters of patients with strokes].

This article is aimed to present a design of a testing device for knee joint kinematics parameters of patients with strokes. We used Auto CAD to design the knee joint sensor device, and applied multifunction data acquisition card to collect sensor data. Then we transferred the data to the computer to quantitatively detect and analyze the joint angle, angular velocity and other parameters. The software system used Visual C+ + MFC frame and shared database Access, and used ADO technology to realize the collection between the software system and patient information system. After a preliminary test of 20 healthy subjects, the result showed that intraclass correlation coefficient (ICC) was more than 0.8, indicating a good reliability of the instrument.

[Design and realization of training methods based on sit-to-stand balance ability training apparatus].

In this research, a new set of training methods was created and put into practice to train the body's sit-to-stand balancing ability, and controlling ability of body's gravity centre of patients with cerebrovascular diseases. It contains a series of motion trainings, including sitting position training, standing position training, sit-to-stand training and game training, etc. There is also a series of trainings of feedback included in this set of methods, such as internal feedback, external feedback, bandwidth feedback, conclusive feedback, visual feedback, acoustical feedback and so on. The trainings resulted in better controlling ability of body's centre of gravity in the sit-to-stand process of patients after a period of training. It indicated that these targeted-designed trainings, especially the feedback trainings, could effectively help the under-trained patients to have a better knowledge of their current situation, and in turn to have a better knowledge of which aspects they should take more training to achieve a better therapeutic effect.

[An intelligent stimulator based on electromyography feature extraction].

In order to improve the conventional stimulator with non-standard parameters and poor efficacy using the passive mode of the treatment, we developed an intelligent stimulator to combine biofeedback with functional electrical stimulation. Through the non-invasive measurement and feature extraction of electromyography signals, a value of root mean square can be obtained based on delta-sigma computational technique. In the newly designed stimulator, electromyography feature extraction and feedback control were used to intelligently control the rehabilitation treatment. Both bidirectional detection technologies for stimulated current and programmable interactive way using touchscreen were employed so that the treatment parameters can be accurately quantified and set up. The experiment showed that the design requirements were achieved.

Characteristics and molecular mechanism of adhesion proteins on reused hemodialysis membranes.

In order to study the mechanism of protein adhesion on the Fresenius F6 polysulfone membrane dialyzer, two-dimensional gel electrophoresis, LC-ESI-MS/MS and bioinformatics methods were used to analyze the protein which adhered to the dialyzer membrane. Six of the adhered proteins account for more than 50% of the total 179 proteins, i.e. ficolin precursor, complement C3 precursor, 3 variants of MASP1 and albumin. The results also showed that easily adhered proteins have a greater percentage of acidic amino acids (p < 0.01). The isoelectric point of the 20 proteins with the most deposits is 6.2 +/- 1.08, which is obviously lower than of those with the least deposits (7.56 +/- 1.36, p < 0.01). The dipole moment of a polysulfone membrane molecule has a tendency to absorb molecules with a negative charge. These results are of significance in understanding and improving membrane protein interactions.

[Species and characteristics of protein adsorption on reused hemodialysis membranes].

The aim of this study was to systemically analyze the proteins that adsorbed on the surface of hemodialysis membrane. The Fresenius F6 reused polysulfone dialyzers were selected as the research objects. The methodology we used encompassed the digestion of protein in rinsed solution and the separation of peptide mixture in virtue of RP-HPLC followed by ESI-MS/MS identification in orde to get their adsorption behavior, species and characteristics. The results illustrated that, after being rinsed by reverse osmosis (RO) water, 179 species of the protein adsorbed on the hemodialysis membrane, most of which were acidulous and middle or small weight protein molecules. The data from counting the numbers of peptides showed there were 5 species of high-abundant proteins (the contents being above 5% each), namely Ficolin-2 precursor, Complement C3 precursor (Fragment), Mannan-binding lectin serine protease 1 isoform 2 precursor, Complement-activating component of Ra-reactive factor precursor, and Mannan-binding lectin serine protease 1 isoform 3. These proteins are in close relationship with human immune system; moreover, they are of great significance to the clinical mechanism for cleaning reuse hemodialyzers and to the development of new hemodialysis materials.

[Ingredients of membrane adhesion in reused dialyzer].

Selecting reused polysulfone membrane (PSM) dialyzers as research objects, we mainly analyzed quantitatively the adhesion ingredients which obstructed the passage through the membrane, and we investigated the differences of residual contaminants on the surface of PSM in the cases of various reuse times. The results illustrated that after the completion of dialysis, the dialyzer was first cleaned by reverse osmosis (RO) water to remove the protein adsorpted. Then we used 2% sodium hypochlorite (NaClO) solution to soak it, and the glucose, cholesterol and triglyceride adsorpted were dissolved off. Meanwhile, the quantity of most of adsorption gradually increased with the increase of reuse times.