Evolution of the surface temperature of pianists' arm muscles using infrared thermography

Musculoskeletal disorders are very frequent among musicians. Diagnosis is difficult due to the lack of objective tests and the multiplicity of symptoms. Treatment is also problematic and often requires that the musician stop playing. Most of these disorders are inflammatory in nature, and therefore involve temperature changes in the affected regions. Temperature measurements were recorded with an infrared camera. In this paper we present an overview of the temperature measurements made in the arms of 8 pianists during regular piano practice sessions.

An Analysis Of Chirpp Data To Predict Severe ATV Injuries Using Artificial Neural Networks

This paper describes the development of a tool to predict the severity of all-terrain vehicle (ATV) injuries using artificial neural networks (ANNs). The data was obtained from the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP). The main objective of the study was to identify the contribution of input variables in predicting severe injury or death. An ANN architecture with 9 hidden nodes and one hidden layer resulted in optimal performance: a logarithmic-sensitivity index of 0.099, sensitivity of 47.3%, specificity of 80.8%, correct classification rate (CCR) of 68.6% and receiver operating curve (ROC) area of 0.711. The minimum data set that can help predict injury severity is discussed.

Survey of clinical engineering in developing countries and model for technology acquisition and diffusion

An international survey was conducted in 2003 to assess the status of clinical engineering services delivered to hospitals in developing countries. Data was collected from Asia, Africa, Latin America and Mexico. The responses were compared to two previous studies done in industrialized countries. A model of medical technology acquisition and diffusion is presented.

Predicting clinical outcomes for newborns using two artificial intelligence approaches

Two different approaches, based on artificial neural networks (ANN) and fuzzy logic, were used to predict a number of outcomes of newborns: How they would be delivered, their 5 minute Apgar score, and neonatal mortality. The goal was to assess whether the methods would be comparable or whether they would perform differently for different outcomes. The results were comparable for Correct Classification Rate (CCR) and Specificity (true negative cases). Sensitivity (true positive cases) was slightly higher for the back-propagation feed-forward ANN than using the Fuzzy-Logic Classifier (FLC). Since this is one single database and a very large one, it is possible that the FLC would perform better than the ANN for very small databases, as shown by some of the co-authors in the past. The next step will be to test a small database with both methods to assess strengths and weaknesses with the intent to use both if needed with some medical data in the future.

Towards ethical decision support and knowledge management in neonatal intensive care

Recent studies in neonatal medicine, clinical nursing, and cognitive psychology have indicated the need to augment current decision-making practice in neonatal intensive care units with computerized, intelligent decision support systems. Rapid progress in artificial intelligence and knowledge management facilitates the design of collaborative ethical decision-support tools that allow clinicians to provide better support for parents facing inherently difficult choices, such as when to withdraw aggressive treatment. The appropriateness of using computers to support ethical decision-making is critically analyzed through research and literature review. In ethical dilemmas, multiple diverse participants need to communicate and function as a team to select the best treatment plan. In order to do this, physicians require reliable estimations of prognosis, while parents need a highly useable tool to help them assimilate complex medical issues and address their own value system. Our goal is to improve and structuralize the ethical decision-making that has become an inevitable part of modern neonatal care units. The paper contributes to clinical decision support by outlining the needs and basis for ethical decision support and justifying the proposed development efforts.

Automated image segmentation for breast analysis using infrared images

In order to realize a fully automated thermogram analysis package for breast cancer detection, it is necessary to identify the region of interest in the thermal image prior to analysis. A nearly fully automated approach is outlined that is able to successfully locate the breast regions in most of the images analyzed. The approach consists of a sequence of Canny edge detectors to determine the body boundaries and to isolate the most likely candidates for the bottom breast boundary. Three different strategies for identifying the bottom breast boundary are investigated: a variation of the Hough transform to identify the curved edges in the image, an algorithm used to detect the longest connected edges that are not part of the body boundary, and a third approach involving the density of detected edges in the breast region. The last two methods show great promise in successfully segmenting the breasts.

Towards a web services infrastructure for perinatal, obstetrical, and neonatal clinical decision support

This paper presents the design of a unifying infrastructure for clinical decision support systems (CDSSs) and medical data relating to the perinatal life cycle. The diverse CDSSs designed for deployment within the perinatal life cycle to improve care, such as Artificial Neural Networks and Case-Based Reasoners, are integrated using the eXtended Markup Language (XML) and are subsequently offered as a secure web service. These web services are accessible from anywhere within the hospital information system and from remote authorized sites. The goal of such an infrastructure is to provide integrated CDSS processing in a complex distributed environment, in order to support real-time physician decision-making. This design provides a novel web services infrastructure implementation and offers a strong case study for deploying and evaluating the web services paradigm within a health care environment.

Predicting high-risk preterm birth using artificial neural networks

A reengineered approach to the early prediction of preterm birth is presented as a complimentary technique to the current procedure of using costly and invasive clinical testing on high-risk maternal populations. Artificial neural networks (ANNs) are employed as a screening tool for preterm birth on a heterogeneous maternal population; risk estimations use obstetrical variables available to physicians before 23 weeks gestation. The objective was to assess if ANNs have a potential use in obstetrical outcome estimations in low-risk maternal populations. The back-propagation feedforward ANN was trained and tested on cases with eight input variables describing the patient's obstetrical history; the output variables were: 1) preterm birth; 2) high-risk preterm birth; and 3) a refined high-risk preterm birth outcome excluding all cases where resuscitation was delivered in the form of free flow oxygen. Artificial training sets were created to increase the distribution of the underrepresented class to 20%. Training on the refined high-risk preterm birth model increased the network's sensitivity to 54.8%, compared to just over 20% for the nonartificially distributed preterm birth model.

Risk factors for Apgar score using artificial neural networks

Artificial Neural Networks (ANNs) have been used in identifying the risk factors for many medical outcomes. In this paper, the risk factors for low Apgar score are introduced. This is the first time, to our knowledge, that the ANNs are used for Apgar score prediction. The medical domain of interest used is the perinatal database provided by the Perinatal Partnership Program of Eastern and Southeastern Ontario (PPPESO). The ability of the feed forward back propagation ANNs to generate strong predictive model with the most influential variables is tested. Finally, minimal sets of variables (risk factors) that are important in predicting Apgar score outcome without degrading the ANN performance are identified.

Predicting probability of mortality in the neonatal intensive care unit

Artificial neural networks can be trained to predict outcomes in a neonatal intensive care unit (NICU). This paper expands on past research and shows that neural networks trained by the maximum likelihood estimation criterion will approximate the ;a posteriori probability' of NICU mortality. A gradient ascent method for the weight update of three-layer feed-forward neural networks was derived. The neural networks were trained on NICU data and the results were evaluated by performance measurement techniques, such as the Receiver Operating Characteristic Curve and the Hosmer-Lemeshow test. The resulting models applied as mortality prognostic screening tools are presented.

Identifying risk factors for two complication types for neonatal intensive care patients (NICU)

This paper discusses the results of applying artificial neural networks to predicting complication for neonatal intensive care patients. Risk factors that lead to necrotizing entero-colitis or broncho-pulmonary dysplasia were identified. Future work will expand this work to other outcomes and add probability information to the estimations.

Conceptual Framework of Knowledge Management for Ethical Decision-Making Support in Neonatal Intensive Care

This research is built on the belief that artificial intelligence estimations need to be integrated into clinical social context to create value for health-care decisions. In sophisticated neonatal intensive care units (NICUs), decisions to continue or discontinue aggressive treatment are an integral part of clinical practice. High-quality evidence supports clinical decision-making, and a decision-aid tool based on specific outcome information for individual NICU patients will provide significant support for parents and caregivers in making difficult "ethical" treatment decisions. In our approach, information on a newborn patient's likely outcomes is integrated with the physician's interpretation and parents' perspectives into codified knowledge. Context-sensitive content adaptation delivers personalized and customized information to a variety of users, from physicians to parents. The system provides structuralized knowledge translation and exchange between all participants in the decision, facilitating collaborative decision-making that involves parents at every stage on whether to initiate, continue, limit, or terminate intensive care for their infant.

Improvement and automation of artificial neural networks to estimate medical outcomes

The lengthy process of manually optimizing a feedforward backpropagation artificial neural network (ANN) provided the incentive to develop an automated system that could fine-tune the network parameters without user supervision. A new stopping criterion was introduced--the logarithmic-sensitivity index--that manages a good balance between sensitivity and specificity of the output classification. The automated network automatically monitored the classification performance to determine when was the best time to stop training-after no improvement in the performance measure (either highest correct classification rate, lowest mean squared error or highest log-sensitivity index value) occurred in the subsequent 500 epochs. Experiments were performed on three medical databases: an adult intensive care unit, a neonatal intensive care unit and a coronary surgery patient database. The optimal network parameter settings found by the automated system were similar to those found manually. The results showed that the automated networks performed equally well or better than the manually optimized ANNs, and the best classification performance was achieved using the log-sensitivity index as a stopping criterion.

Quantitative assessment of pain-related thermal dysfunction through clinical digital infrared thermal imaging

Background

The skin temperature distribution of a healthy human body exhibits a contralateral symmetry. Some nociceptive and most neuropathic pain pathologies are associated with an alteration of the thermal distribution of the human body. Since the dissipation of heat through the skin occurs for the most part in the form of infrared radiation, infrared thermography is the method of choice to study the physiology of thermoregulation and the thermal dysfunction associated with pain. Assessing thermograms is a complex and subjective task that can be greatly facilitated by computerised techniques.

Methods

This paper presents techniques for automated computerised assessment of thermal images of pain, in order to facilitate the physician's decision making. First, the thermal images are pre-processed to reduce the noise introduced during the initial acquisition and to extract the irrelevant background. Then, potential regions of interest are identified using fixed dermatomal subdivisions of the body, isothermal analysis and segmentation techniques. Finally, we assess the degree of asymmetry between contralateral regions of interest using statistical computations and distance measures between comparable regions.

Results

The wavelet domain-based Poisson noise removal techniques compared favourably against Wiener and other wavelet-based denoising methods, when qualitative criteria were used. It was shown to improve slightly the subsequent analysis. The automated background removal technique based on thresholding and morphological operations was successful for both noisy and denoised images with a correct removal rate of 85% of the images in the database. The automation of the regions of interest (ROIs) delimitation process was achieved successfully for images with a good contralateral symmetry. Isothermal division complemented well the fixed ROIs division based on dermatomes, giving a more accurate map of potentially abnormal regions. The measure of distance between histograms of comparable ROIs allowed us to increase the sensitivity and specificity rate for the classification of 24 images of pain patients when compared to common statistical comparisons.

Conclusions

We developed a complete set of automated techniques for the computerised assessment of thermal images to assess pain-related thermal dysfunction.

Weight-elimination neural networks applied to coronary surgery mortality prediction

The objective was to assess the effectiveness of the weight-elimination cost function in improving classification performance of artificial neural networks (ANNs) and to observe how changing the a priori distribution of the training set affects network performance. Backpropagation feedforward ANNs with and without weight-elimination estimated mortality for coronary artery surgery patients. The ANNs were trained and tested on cases with 32 input variables describing the patient's medical history; the output variable was in-hospital mortality (mortality rates: training 3.7%, test 3.8%). Artificial training sets with mortality rates of 20%, 50%, and 80% were created to observe the impact of training with a higher-than-normal prevalence. When the results were averaged, weight-elimination networks achieved higher sensitivity rates than those without weight-elimination. Networks trained on higher-than-normal prevalence achieved higher sensitivity rates at the cost of lower specificity and correct classification. The weight-elimination cost function can improve the classification performance when the network is trained with a higher-than-normal prevalence. A network trained with a moderately high artificial mortality rate (artificial mortality rate of 20%) can improve the sensitivity of the model without significantly affecting other aspects of the model's performance. The ANN mortality model achieved comparable performance as additive and statistical models for coronary surgery mortality estimation in the literature.

Extending ventilation duration estimations approach from adult to neonatal intensive care patients using artificial neural networks

In earlier work, the research group successfully used artificial neural networks (ANNs) to estimate ventilation duration for adult intensive care unit (ICU) patients. The ANNs performed well in terms of correct classification rate (CCR) and average squared error (ASE) classifying the outcome into two classes: whether patients were ventilated for less than/equal to or for more than 8 h (< or >). The objective of new work was to apply this adult model to the estimation of ventilation with neonatal ICU (NICU) patient records. The performance obtained with the neonatal patients was comparable to that previously found with the adult database, again as measured in terms of a maximum CCR and a minimum ASE. The effectiveness of using the weight-elimination technique in controlling overfitting was again validated for the neonatal patients as it had been for our adult patients. It was concluded that the approach developed for ICU adult patients was also successfully applied to a different medical environment: neonatal ICU patients.

Influence of missing values on artificial neural network performance

The problem of databases containing missing values is a common one in the medical environment. Researchers must find a way to incorporate the incomplete data into the data set to use those cases in their experiments. Artificial neural networks (ANNs) cannot interpret missing values, and when a database is highly skewed, ANNs have difficulty identifying the factors leading to a rare outcome. This study investigates the impact on ANN performance when predicting neonatal mortality of increasing the number of cases with missing values in the data sets. Although previous work using the Canadian Neonatal Intensive Care Unit (NICU) Network s database showed that the ANN could not correctly classify any patients who died when the missing values were replaced with normal or mean values, this problem did not arise as expected in this study. Instead, the ANN consistently performed better than the constant predictor (which classifies all cases as belonging to the outcome with the highest training set a priori probability) with a 0.6-1.3% improvement over the constant predictor. The sensitivity of the models ranged from 14.5-20.3% and the specificity ranged from 99.2- 99.7%. These results indicate that nearly 1 in 5 babies who will eventually die are correctly classified by the ANN, and very few babies were incorrectly identified as patients who will die. These findings are important for patient care, counselling of parents and resource allocation.

Clinical decision support systems for intensive care units: using artificial neural networks

The paper provides an overview of applications of artificial neural networks (ANNs) to various medical problems, with a particular focus on the intensive care unit environment (ICU). Several technical approaches were tested to see whether they improve the ANN performance in estimating medical outcomes and resource utilization in adult ICUs. These experiments include: (1) use of the weight-elimination cost function; (2) use of 'high' and 'low' nodes for input variables; (3) verifying the effect of the total number of input variables on the results; (4) testing the impact of the value of the constant predictor on the performance of the ANNs. The developments presented intend to help medical and nursing personnel to assess patient status, assist in making a diagnosis, and facilitate the selection of a course of therapy.

Clinical decision-support systems for intensive care units using case-based reasoning

The artificial intelligence approach used in this work focusses on case-based reasoning techniques for the estimation of medical outcomes and resource utilization. The systems were designed with a view to help medical and nursing personnel to assess patient status, assist in making a diagnosis, and facilitate the selection of a course of therapy. The initial prototype provided information on the closest-matching patient cases to the newest patient admission in an adult intensive care unit (ICU). The system was subsequently re-designed for use in a neonatal ICU. The results of a short clinical pilot evaluation performed in both adult and neonatal units are reported and have led to substantial improvement of the prototype. Future work will include longer-term clinical trials for both adult and neonatal ICUs, once all the software changes have been made to both prototypes in response to the comments of the users made during the preliminary evaluations. To date, the results are very encouraging and physician interest in the potential clinical usefulness of these two systems remains high, and particularly so in the new testing environment in Ottawa.

Decision-support and intelligent tutoring systems in medical education

One of the challenges in medical education is to teach the decision-making process. This learning process varies according to the experience of the student and can be supported by various tools. In this paper we present several approaches that can strengthen this mechanism, from decision-support tools, such as scoring systems, Bayesian models, neural networks, to cognitive models that can reproduce how the students progressively build their knowledge into memory and foster pedagogic methods.

Selective sampling to overcome skewed a priori probabilities with neural networks

Highly skewed a priori probabilities present challenges for researchers developing medical decision aids due to a lack of information on the rare outcome of interest. This paper attempts to overcome this obstacle by artificially increasing the mortality rate of the training sets. A weight pruning technique called weight-elimination is also applied to this coronary artery bypass grafting (CABG) database to assess its impact on the artificial neural network's (ANN) performance. The results showed that increasing the mortality rate improved the sensitivity rates at the cost of the other performance measures, and the weight-elimination cost function improved the sensitivity rate without seriously affecting the other performance measures.