Neural network modeling for surgical decisions on traumatic brain injury patients

Computerized medical decision support systems have been a major research topic in recent years. Intelligent computer programs were implemented to aid physicians and other medical professionals in making difficult medical decisions. This report compares three different mathematical models for building a traumatic brain injury (TBI) medical decision support system (MDSS). These models were developed based on a large TBI patient database. This MDSS accepts a set of patient data such as the types of skull fracture, Glasgow Coma Scale (GCS), episode of convulsion and return the chance that a neurosurgeon would recommend an open-skull surgery for this patient. The three mathematical models described in this report including a logistic regression model, a multi-layer perceptron (MLP) neural network and a radial-basis-function (RBF) neural network. From the 12,640 patients selected from the database. A randomly drawn 9480 cases were used as the training group to develop/train our models. The other 3160 cases were in the validation group which we used to evaluate the performance of these models. We used sensitivity, specificity, areas under receiver-operating characteristics (ROC) curve and calibration curves as the indicator of how accurate these models are in predicting a neurosurgeon's decision on open-skull surgery. The results showed that, assuming equal importance of sensitivity and specificity, the logistic regression model had a (sensitivity, specificity) of (73%, 68%), compared to (80%, 80%) from the RBF model and (88%, 80%) from the MLP model. The resultant areas under ROC curve for logistic regression, RBF and MLP neural networks are 0.761, 0.880 and 0.897, respectively (P < 0.05). Among these models, the logistic regression has noticeably poorer calibration. This study demonstrated the feasibility of applying neural networks as the mechanism for TBI decision support systems based on clinical databases. The results also suggest that neural networks may be a better solution for complex, non-linear medical decision support systems than conventional statistical techniques such as logistic regression.     

Comparison of hospital charge prediction models for colorectal cancer patients: neural network vs. decision tree models

Analysis and prediction of the care charges related to colorectal cancer in Korea are important for the allocation of medical resources and the establishment of medical policies because the incidence and the hospital charges for colorectal cancer are rapidly increasing. But the previous studies based on statistical analysis to predict the hospital charges for patients did not show satisfactory results. Recently, data mining emerges as a new technique to extract knowledge from the huge and diverse medical data. Thus, we built models using data mining techniques to predict hospital charge for the patients. A total of 1,022 admission records with 154 variables of 492 patients were used to build prediction models who had been treated from 1999 to 2002 in the Kyung Hee University Hospital. We built an artificial neural network (ANN) model and a classification and regression tree (CART) model, and compared their prediction accuracy. Linear correlation coefficients were high in both models and the mean absolute errors were similar. But ANN models showed a better linear correlation than CART model (0.813 vs. 0.713 for the hospital charge paid by insurance and 0.746 vs. 0.720 for the hospital charge paid by patients). We suggest that ANN model has a better performance to predict charges of colorectal cancer patients.     

An evaluation of intelligent prognostic systems for colorectal cancer

In this paper we describe attempts at building a robust model for predicting the length of survival of patients with colorectal cancer. The aim of the research, reported in this paper, is to study the effective utilisation of artificial intelligence techniques in the medical domain. We suggest that an important research objective of proponents of intelligent prognostic systems must be to evaluate the additionality that AI techniques can bring to an already well-established field of medical prognosis. Towards this end, we compare a number of different AI techniques that lend themselves to the task of predicting survival in colorectal cancer patients. We describe the pros and cons of each of these methods using the usual metrics of accuracy and perspicuity. We then present the notion of intelligent hybrid systems and evaluate the role that they may potentially play in developing robust prognostic models. In particular we evaluate a hybrid system that utilises the k Nearest Neighbour technique in conjunction with Genetic Algorithms. We describe a number of innovations used within this hybrid paradigm used to build the prognostic model. We discuss the issue of censored patients and how this issue can be tackled within the various models used. In keeping with our objective of studying the additionality that AI techniques bring to building prognostic models, we use Cox's regression as a standard and compare each AI technique with it, attempting to discover their capabilities in enhancing prognostic methods in medicine. In doing so we address two main questions--which model fits the data best?, and are the results obtained by the various AI techniques significantly different from those of Cox's regression? We conclude this paper by discussing future enhancements to the work presented and lessons learned from the study to date.     

Important changes in medical x-ray imaging facility shielding design methodology. A brief summary of recommendations in NCRP Report No. 147

The recently published Report No. 147 of The National Council on Radiation Protection and Measurements entitled "Structural shielding design for medical x-ray imaging facilities" provides an update of shielding recommendations for x rays used for medical imaging. The goal of this report is to ensure that the shielding in these facilities limits radiation exposures to employees and members of the public to acceptable levels. Board certified medical and health physicists, as defined in this report, are the "qualified experts" who are competent to design radiation shielding for these facilities. As such, physicists must be aware of the new technical information and the changes from previous reports that Report No. 147 supersedes. In this article we summarize the new data, models and recommendations for the design of radiation barriers in medical imaging facilities that are presented in Report No. 147.     

基于人工智能和临床诊断的上肢康复评估方法研究进展

脑卒中导致患者肢体运动功能障碍或缺失,严重影响了患者的生活质量。在上肢康复治疗中,医生需要对患者的上肢进行主观康复评估,但这种方法误差大、成本高。因此,人们将人工智能技术应用到医疗康复领域。本文总结了基于s EMG信号特征、运动轨迹误差特征、关节运动角度特征、关节角速度特征的客观评估方法,以及Brunnstrom等级评价法、上田敏评价法、Fugl-Meyer量表评价法、Wolf运动功能测试评价方法等主观评估方法。最后,本文认为现有的客观评估方法普遍受到训练资料过少、特征单一等因素影响。主观评估方法普遍受到评估时间过长、易受主观影响等因素影响。未来的客观评估算法还应在算法准确性、训练资料规模、多特征融合等方面继续改进。     

Machine learning of clinical performance in a pancreatic cancer database

ObjectiveWe consider predictive models for clinical performance of pancreatic cancer patients based on machine learning techniques. The predictive performance of machine learning is compared with that of the linear and logistic regression techniques that dominate the medical oncology literature.Methods and materialsWe construct predictive models over a clinical database that we have developed for the University of Massachusetts Memorial Hospital in Worcester, Massachusetts, USA. The database contains retrospective records of 91 patient treatments for pancreatic tumors. Classification and regression targets include patient survival time, Eastern Cooperative Oncology Group (ECOG) quality of life scores, surgical outcomes, and tumor characteristics. The predictive performance of several techniques is described, and specific models are presented.ResultsWe show that machine learning techniques attain a predictive performance that is as good as, or better than, that of linear and logistic regression, for target attributes that include tumor N and T stage, survival time, and ECOG quality of life scores. Bayesian techniques are found to provide the best performance overall. For tumor size as the target attribute, however, logistic regression (respectively linear regression in the case of a numerical as opposed to discrete target) performs best. Preprocessing in the form of attribute selection and supervised attribute discretization improves predictive performance for most of the predictive techniques and target attributes considered.ConclusionMachine learning provides techniques for improved prediction of clinical performance. These techniques therefore merit consideration as valuable alternatives to traditional multivariate regression techniques in clinical medical studies.     

Comparison of artificial neural network and logistic regression models for prediction of mortality in head trauma based on initial clinical data

Background  

In recent years, outcome prediction models using artificial neural network and multivariable logistic regression analysis have been developed in many areas of health care research. Both these methods have advantages and disadvantages. In this study we have compared the performance of artificial neural network and multivariable logistic regression models, in prediction of outcomes in head trauma and studied the reproducibility of the findings.     

Modeling medical prognosis: survival analysis techniques

Medical prognosis has played an increasing role in health care. Reliable prognostic models that are based on survival analysis techniques have been recently applied to a variety of domains, with varying degrees of success. In this article, we review some methods commonly used to model time-oriented data, such as Kaplan-Meier curves, Cox proportional hazards, and logistic regression, and discuss their applications in medical prognosis. Nonlinear, nonparametric models such as neural networks have increasingly been used for building prognostic models. We review their use in several medical domains and discuss different implementation strategies. Advantages and disadvantages of these methods are outlined, as well as pointers to pertinent literature.     

兔椎间盘退变与修复模型的研究现状*

椎间盘退变模型是研究椎间盘退变疾病的基础和关键之一。兔退变椎间盘模型具有操作简单、可重复性好等特点被国内外学者广泛应用。兔椎间盘退变模型包括体内模型、体外模型等。体内模型根据损伤类别包括:机械损伤模型、化学损伤模型、异常应力模型、脊柱不稳模型、脊柱融合模型等;体外模型包括椎间盘细胞模型、椎间盘组织模型等。本文根据近年兔腰椎间盘各种退变模型与修复的研究现状与进展作一综述。     

Computerized assessment of motion-contaminated calcified plaques in cardiac multidetector CT

An automated method for evaluating the image quality of calcified plaques with respect to motion artifacts in noncontrast-enhanced cardiac computed tomography (CT) images is introduced. This method involves using linear regression (LR) and artificial neural network (ANN) regression models for predicting two patient-specific, region-of-interest-specific, reconstruction-specific and temporal phase-specific image quality indices. The first is a plaque motion index, which is derived from the actual trajectory of the calcified plaque and is represented on a continuous scale. The second is an assessability index, which reflects the degree to which a calcified plaque is affected by motion artifacts, and is represented on an ordinal five-point scale. Two sets of assessability indices were provided independently by two radiologists experienced in evaluating cardiac CT images. Inputs for the regression models were selected from 12 features characterizing the dynamic, morphological, and intensity-based properties of the calcified plaques. Whereas LR-velocity (LR-V) used only a single feature (three-dimensional velocity), the LR-multiple (LR-M) and ANN regression models used the same subset of these 12 features selected through stepwise regression. The regression models were parameterized and evaluated using a database of simulated calcified plaque images from the dynamic NCAT phantom involving nine heart rate/multi-sector gating combinations and 40 cardiac phases covering two cardiac cycles. Six calcified plaques were used for the plaque motion indices and three calcified plaques were used for both sets of assessability indices. In one configuration, images from the second cardiac cycle were used for feature selection and regression model parameterization, whereas images from the first cardiac cycle were used for testing. With this configuration, repeated measures concordance correlation coefficients (CCCs) and associated 95% confidence intervals for the LR-V, LR-M, and ANN were 0.817 [0.785, 0.848], 0.894 [0.869, 0.916], and 0.917 [0.892, 0.936] for the plaque motion indices. For the two sets of assess-ability indices, CCC values for the ANN model were 0.843 [0.791, 0.877] and 0.793 [0.747, 0.828]. These two CCC values were statistically greater than the CCC value of 0.689 [0.648, 0.727], which was obtained by comparing the two sets of assessability indices with each other. These preliminary results suggest that the variabilities of assessability indices provided by regression models can lie within the variabilities of the indices assigned by independent observers. Thus, the potential exists for using regression models and assessability indices for determining optimal phases for cardiac CT image interpretation.     

Simple linear regression in medical research

This article discusses the method of fitting a straight line to data by linear regression and focuses on examples from 36 Original Articles published in the Journal in 1978 and 1979. Medical authors generally use linear regression to summarize the data (as in 12 of 36 articles in my survey) or to calculate the correlation between two variables (21 of 36 articles). Investigators need to become better acquainted with residual plots, which give insight into how well the fitted line models the data, and with confidence bounds for regression lines. Statistical computing packages enable investigators to use these techniques easily.     

Integrating regression formulas and kernel functions into locally adaptive knowledge-based neural networks: a case study on renal function evaluation

OBJECTIVE: In many medical areas, there exist different regression formulas to predict/evaluate a medical outcome on the same problem, each of them being efficient only in a particular sub-space of the problem space. The paper aims at the development of a generic, incremental learning model that includes all available regression formulas for a particular prediction problem to define local areas of the problem space with their best performing formula along with useful explanation rules. Another objective of the paper is to develop a specific model for renal function evaluation using nine existing formulas. METHODS AND MATERIALS: We have used a connectionist neuro-fuzzy approach and have developed a knowledge-based neural network model (KBNN) which incorporates and adapts incrementally several existing regression formulas and kernel functions. The model incorporates different non-linear regression functions as neurons in its hidden layer and adapts these functions through incremental learning from data in particular local areas of the space. More specifically, each hidden neural node has a pair of functions associated with it--one regression formula, that represents existing knowledge and one Gaussian kernel function, that defines the sub-space of the whole problem space, in which the formula is locally adapted to new data. All these functions are aggregated and changed through incremental learning. The proposed KBNN model is illustrated using a medical dataset of observed patient glomerular filtration rate (GFR) measurements for renal function evaluation. In this case study, the regression function for each cluster is selected by the model from nine formulas commonly used by medical practitioners to predict GFR. 441 GFR data vectors from 141 patients taken from 12 sites in Australia and New Zealand have been used as a case study experimental data set. RESULTS: The proposed GFR prediction model, based on the proposed generic KBNN model, outperforms at least by 10% accuracy any of the individual regression formulas or a standard neural network model. Furthermore, we have derived locally adapted regression formulas to perform best on local clusters of data along with useful explanatory rules. CONCLUSION: The proposed KBNN model manifests better accuracy then existing regression formulas or neural network models for renal function evaluation and extracts modified formulas that perform well in local areas of the problem space.     

Introduction to spiking neural networks: Information processing, learning and applications

The concept that neural information is encoded in the firing rate of neurons has been the dominant paradigm in neurobiology for many years. This paradigm has also been adopted by the theory of artificial neural networks. Recent physiological experiments demonstrate, however, that in many parts of the nervous system, neural code is founded on the timing of individual action potentials. This finding has given rise to the emergence of a new class of neural models, called spiking neural networks. In this paper we summarize basic properties of spiking neurons and spiking networks. Our focus is, specifically, on models of spike-based information coding, synaptic plasticity and learning. We also survey real-life applications of spiking models. The paper is meant to be an introduction to spiking neural networks for scientists from various disciplines interested in spike-based neural processing.     

Statistical modelling of general practice medicine for computer assisted data entry in electronic medical record systems

Electronic medical record (EMR) systems have much potential, however, there are still a number of issues that need to be resolved before EMRs are widely accepted. One of these issues is the data input task, a potentially serious practical barrier to on-line medical computer usage. This paper reports the empirical modelling of data input requirements for physicians who use a problem-orientated medical record system. Three statistical models (Bayesian conditional probability, multiple linear regression and discriminant analysis) to predict drug treatment given problem diagnoses are derived from EMRs of 2500 general Practice encounters. Two metrics are used to measure the predictive power of the models considering both the number of drugs correctly predicted and the strength with which the models predict them. The models are tested on 500 unseen records from the same patient-physician population and the data used to build the models. The Bayesian model produces the best predictions on unseen data and is also the easiest model to compute. A prototype interface that enables new patient cases to be entered is constructed to demonstrate how the predictive power of the model can translate into benefits in the data entry task.     

The role of depression in the quality of life among patients undergoing renal substitutive therapy

BACKGROUND: Depression is highly prevalent among patients with end-stage renal disease, nevertheless few patients are assessed or offered medical treatment to minimize its effects. This study assessed quality of life among these patients and studied the association between end-stage renal disease and depression. MATERIAL AND METHODS: We carried out a cross-sectional study with 123 patients over 19 who were undergoing renal substitutive therapy. Quality of life and depression were assessed using the Kidney Disease Quality of Life Short Form-36 and the Beck Depression Inventor. In order to measure the patients' metabolic state, we carried out medical and laboratory tests. Quality of life predictors were analyzed with multiple ordinal logistic regression models. RESULTS: The highest scores from the generic core belonged to social functioning dimensions (62.7) and mental health (65.9). For the specific core, the highest scores were in dimensions associated with support offered by the dialysis team (78.2) and from social support networks (75.3). Depression was the most consistent predictor of quality of life. CONCLUSION: The use of programs and measuring tools to measure quality of life prior to and during renal dialysis or hemodialysis, as well as a timely psychiatric evaluation, can be very useful in monitoring improvement, decline and response to anti-depressant treatment throughout the course of end-stage renal disease.     

医学图像配准技术及其研究进展

目的:对近年来的医学图像配准技术及其研究进展情况进行详尽地综述和讨论,从而为开展医学图像配准技术在医学图像三维重建、医学图像可视化和定量分析方面的研究提供参考.方法:首先,查阅国内外近年来医学图像配准技术研究的权威文献;然后,深入分析和研究这些文献所介绍方法的特点、存在的问题,并针对存在的问题提出可能的解决方案.结果:通过对近年来医学图像配准算法的最新研究进展情况进行深人细致地分析和讨论,在比较了一些典型算法的特点及其应用的基础上,对医学图像配准技术的发展进行了展望.结论:使用最优化策略改进图像配准质量以及对非刚体图像配准的研究是今后医学图像配准的发展方向.     

Modeling molecular computing systems by an artificial chemistry - its expressive power and application

Artificial chemistries are mainly used to construct virtual systems that are expected to show behavior similar to living systems. In this study, we explore possibilities of applying an artificial chemistry to modeling natural biochemical systems-or, to be specific, molecular computing systems-and show that it may be a useful modeling tool for molecular computation. We previously proposed an artificial chemistry based on string pattern matching and recombination. This article first demonstrates that this artificial chemistry is computationally universal if it has only rules that have one reactant or two reactants. We think this is a good property of an artificial chemistry that models molecular computing, because natural elementary chemical reactions, on which molecular computing is based, are mostly unimolecular or bimolecular. Then we give two illustrative example models for DNA computing in our artificial chemistry: one is for the type of computation called the Adleman-Lipton paradigm, and the other is for a DNA implementation of a finite automaton. Through the construction of these models we observe preferred properties of the artificial chemistry for modeling molecular computing, such as having no spatial structure and being flexible in choosing levels of abstraction.     

Restoration of the pubourethral ligament with platelet rich plasma for the treatment of stress urinary incontinence

Stress urinary incontinence (SUI) is a major health problem, which affects nearly 20% of adult women and has a detrimental impact on their daily activities and quality of life. Several surgical techniques have been proposed for the treatment of SUI including the Burch colposuspension, retropubic mid-urethral slings (TVT), trans-obturator tapes (TOT), trans-obturator tapes inside out (TVT-O), bladder neck injections and the insertion of an artificial urethral sphincter. All of these treatments aim to either restore the urethral support, which is naturally preserved by the pubourethral ligament (PUL) or to increase the urethral resistance at rest. Most surgical techniques are associated with a variety of intraoperative and postoperative complications. Platelet rich plasma (PRP) is extremely rich in growth factors and cytokines, which regulate tissue reconstruction and has been studied extensively among trauma patients and trauma experimental models. To date, however, there is no evidence to support or oppose its use in women who suffer from SUI due to PUL damage. PRP is an easily produced and relatively inexpensive biologic material. It is produced directly from the patient’s blood and is, thus, superior to synthetic materials in terms of potential adverse effects such as from foreign body reaction. In the present article we summarize the existing evidence in the field, which supports the conduct of animal experimental and clinical studies to elucidate the potential role of PRP in treating SUI.     

Model selection for a medical diagnostic decision support system: a breast cancer detection case

There are a number of different quantitative models that can be used in a medical diagnostic decision support system (MDSS) including parametric methods (linear discriminant analysis or logistic regression), non-parametric models (K nearest neighbor, or kernel density) and several neural network models. The complexity of the diagnostic task is thought to be one of the prime determinants of model selection. Unfortunately, there is no theory available to guide model selection. Practitioners are left to either choose a favorite model or to test a small subset using cross validation methods. This paper illustrates the use of a self-organizing map (SOM) to guide model selection for a breast cancer MDSS. The topological ordering properties of the SOM are used to define targets for an ideal accuracy level similar to a Bayes optimal level. These targets can then be used in model selection, variable reduction, parameter determination, and to assess the adequacy of the clinical measurement system. These ideas are applied to a successful model selection for a real-world breast cancer database. Diagnostic accuracy results are reported for individual models, for ensembles of neural networks, and for stacked predictors.     

Parthenogenesis as an Approach to Pluripotency: Advantages and Limitations Involved

Embryonic stem cells (ESCs) are invaluable cells derived from the inner cell mass of the mammalian blastocyst. They have nearly indefinite self-renewal, retain their developmental potential after prolonged periods in culture and display great plasticity that allow them to differentiate into all cell types of the body. They provide exciting opportunities to develop unique models for developmental research and hold great potential for cell and tissue replacement therapy. However, these unique cells cannot be obtained without destroying an embryo and, despite the potential therapeutic usefulness, their derivation in the human raises substantial ethical as well as legal and political concerns because it unavoidably involves the destruction of viable embryos. In the recent years a number of scientific proposals that do not require the generation and subsequent destruction of human embryos have been put forward in an attempt to fill the gap between ethical questions and potential scientific and medical benefits. In this review we briefly summarize data obtained from the literature related to these different alternative approaches and focus in more details on our experience in the derivation of parthenothes, as a possible alternative source for pluripotent cells, discussing the advantages as well as the limits of these cell lines.