SEARCH · Search PubMed
Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
NEWCHEM, an artificial intelligence system for the control of cancer cell growth, is described. This system takes into account the most recent advances in molecular and cellular biology and in cell-drug interaction, and aims to develop optimal strategies for the selective control of cancer cell through qualitative reasoning from first principles at cellular level.
This article presents a further description on the background, significance, and structure characteristics of Quantitative Medicine Simulation and Operation by Computer (QMSOC). Also some basic operators were recommended for calculations of biomedical events such as estimation of substance concentrations, exploration of etiology, evaluation of biomedical effects, etc. At last some differences of QMSOC from other artificial intelligent systems in the medical field were discussed.
PURPOSE: Installation of airbags has reduced the rate of fatal injuries in severe automobile accidents. We report, however, severe ocular injuries in a minor accident as the result of an airbag. CASE REPORT: A front passenger suffered a blunt ocular trauma of her right eye during a collision. The approaching speed was about 31 km/h. The maximum change of velocity in direction of the impact was 19 km/h. Color traces were found on the upper rim of the airbag, apparently from the patient's eye shadow. RESULTS: In the emergency room, visual acuity was reduced to light perception. There was endothelial contusion, traumatic mydriasis, and lens subluxation. A sclopetarian retinopathy developed with a chorioretinal scar. Eight months after the accident visual acuity remained at light projection only. CONCLUSIONS: The eye injuries had very probably been caused by the deploying airbag. Improvements are a better geometry of deployment (e.g., tethered airbags), release at higher impacts only, and "intelligent systems" with additional sensors to avoid potentially hazardous airbag inflation in minor accidents.
Mixture of experts (ME) is a modular neural network architecture for supervised learning. This paper illustrates the use of ME network structure to guide diagnosing of breast cancer. Expectation-maximization (EM) algorithm was used for training the ME so that the learning process is decoupled in a manner that fits well with the modular structure. Diagnosis tasks are among the most interesting activities in which to implement intelligent systems. Specifically, diagnosis is an attempt to accurately forecast the outcome of a specific situation, using as input information obtained from a concrete set of variables that potentially describe the situation. The ME network structure was implemented for breast cancer diagnosis using the attributes of each record in the Wisconsin breast cancer database. To improve diagnostic accuracy, the outputs of expert networks were combined by a gating network simultaneously trained in order to stochastically select the expert that is performing the best at solving the problem. For the Wisconsin breast cancer diagnosis problem, the obtained total classification accuracy by the ME network structure was 98.85%. The ME network structure achieved accuracy rates which were higher than that of the stand-alone neural network models.
The test of the association between dietary intake of specific carotenoids and disease incidence requires the availability of accurate and current food composition data for individual carotenoids. To generate a carotenoid database, an artificial intelligence system was developed to evaluate data for carotenoid content of food in five general categories, namely, number of samples, analytic method, sample handling, sampling plan, and analytic quality control. Within these categories, criteria have been created to rate analytic data for beta-carotene, alpha-carotene, lutein, lycopene, and beta-cryptoxanthin in fruits and vegetables. These carotenoids are also found in human blood. Following the evaluation of data, acceptable values for each carotenoid in the foods were combined to generate a database of 120 foods. The database includes the food description; median, minimum, and maximum values for the specific carotenoids in each food; the number of acceptable values and their references; and a confidence code, which is an indicator of the reliability of a specific carotenoid value for a food. The carotenoid database can be used to estimate the intake of specific carotenoids in order to examine the association between dietary carotenoids and disease incidence.
Two procedures for predicting the carcinogenicity of chemicals are described. One of these (CASE) is a self-learning artificial intelligence system that automatically recognizes activating and/or deactivating structural subunits of candidate chemicals and uses this to determine the probability that the test chemical is or is not a carcinogen. If the chemical is predicted to be carcinogen, CASE also projects its probable potency. The second procedure (CPBS) uses Bayesian decision theory to predict the potential carcinogenicity of chemicals based upon the results of batteries of short-term assays. CPBS is useful even if the test results are mixed (i.e. both positive and negative responses are obtained in different genotoxic assays). CPBS can also be used to identify highly predictive as well as cost-effective batteries of assays. For illustrative purposes the ability of CASE and CPBS to predict the carcinogenicity of a carcinogenic and a non-carcinogenic polycyclic aromatic hydrocarbon is shown. The potential for using the two methods in tandem to increase reliability and decrease cost is presented.
Recently Goldring et al. [Mutation Res., 187 (1987) 67-77] reported the synthesis and purification of a series of nitro-substituted cyclopenta-fused polycyclic aromatic hydrocarbons. On the basis of expected charge distributions, these chemicals were predicted to be potent mutagens and, yet, contrary to expectation, they were found to be only weakly mutagenic for Salmonella. In their discussion, the authors suggest that application of CASE, an artificial intelligence system recently developed in these laboratories, would also not predict the low mutagenicity of this group of chemicals. In the present report, it is shown that CASE, in fact, correctly predicts the low mutagenicity of nitro-substituted cyclopenta-fused polycyclic aromatic hydrocarbons.
An important research task of the EuroMISE Centre is the applied research in the field of electronic health record (EHR) design including electronic medical guidelines and intelligent systems for data mining and decision support. The research in this field was inspired by several European projects. We have proposed a mathematical meta-description of a flexible information storage model based on the experience gathered in cooperation in those projects. In this model, we use two basic structures called a knowledge base and data files. We describe those two structures using the graph theory concepts. Furthermore, we use logical formulas to express conditions that should be valid. Additionally, we present a description of a global system architecture of a 3-tier EHR application with interfaces based on the latest technologies; predominately on Web Services, SOAP, XML, HTTP, CORBA, etc. According to our experience and test results gained from the MUDR EHR usage, we describe an open universal solution, which can be applied as the EHR kernel of hospital information systems. To realize this approach in a daily practice for health professionals we have started a co-operative project with clinical information systems developers. Within that project we are developing a new system for continual shared health care.
Novel drug delivery and biosensing devices have the potential to increase the efficacy of drug therapy by providing physicians and patients the ability to precisely control key therapy parameters. Such "intelligent" systems can enable control of dose amount and the time, rate, and location of drug delivery. We have developed and demonstrated the operation of an electrothermal mechanism to precisely control the delivery of drugs and exposure of biosensors. These microchip devices contain an array of individually sealed and actuated reservoirs, each capped by a thin metal membrane comprised of either gold or multiple layers of titanium and platinum. The passage of a threshold level of electric current through the membrane causes it to disintegrate, thereby exposing the protected contents (drugs or biosensors) of the reservoir to the surrounding environment. This paper describes the theory and experimental characterization of the electrothermal method and includes in vitro release results for a model compound.
Visual motion provides useful information to understand the dynamics of a scene to allow intelligent systems interact with their environment. Motion computation is usually restricted by real time requirements that need the design and implementation of specific hardware architectures. In this paper, the design of hardware architecture for a bio-inspired neural model for motion estimation is presented. The motion estimation is based on a strongly localized bio-inspired connectionist model with a particular adaptation of spatio-temporal Gabor-like filtering. The architecture is constituted by three main modules that perform spatial, temporal, and excitatory-inhibitory connectionist processing. The biomimetic architecture is modeled, simulated and validated in VHDL. The synthesis results on a Field Programmable Gate Array (FPGA) device show the potential achievement of real-time performance at an affordable silicon area.
In this paper (An abbreviated version of some portions of this article appeared in reference Menard and Frezza-Buet (Menard, O., & Frezza-Buet, H. (2004). Rewarded multi-modal neuronal self-organization: Example of the arm reaching movement. In: Proceedings of international conference on advances in intelligent systems theory and application.), as part of the IJCNN 2005 conference proceedings, published under the IEEE copyright.), an original self-organizing model is presented, with experiments highlighting its ability to be used in different frameworks, as phonetic coding dependent on semantics and arm-reaching. The model relies on the coupling of the learning processes that stand at different self-organizing modules, and exhibits dynamics that can be discussed in terms of the binding of different modalities, scattered over the different modules. Such a binding property is based on an emerging constraint of keeping consistency between the modules. This process is induced by partial connectivity and appropriate neural field competition mechanisms.
The past decade has seen some impressive successes in the metabolic engineering of biotechnologically important plant pathways. However, plant metabolic engineering currently proceeds more by trial and error than by intelligent system design. A change in philosophy away from studying pathways in isolation and towards studying metabolism as a network is necessary. To support this development, improvements in technologies for metabolic analysis, a wider adoption of metabolite-profiling approaches and significant innovations in data analysis methodologies are required.
Devices constructed for potential use as invasive bioprobes incorporate a selective receiving site for molecular or ionic recognition, and a transducer which is capable of translating a perturbation of physical chemistry of the determinant-site reaction (interaction) into a usable signal. Four types are envisioned--implants for general hospital use, transient-use probes to replace classical blood tests, short-term implantable probes and the long-term variety. Performance criteria are selectivity, sensitivity, fast response, site-reversible, small, rugged, inexpensive, biocompatible, calibratible, facile use by non-expert personnel and ease of telemetry. These demands, not surprisingly, create enormous challenges to the sensor specialist. With respect to biocompatibility the sensor must not be involved in infection, clot formation or antigenic response, and, furthermore, protein adsorption, etc., which can affect the sensor response should be avoided. Calibration remains a problem of monumental proportions. Many devices drift from calibrated levels even in in vitro experiments, let alone in the implanted milieu. One solution has been to carry out on-line switching between patient blood and standard solutions. However, this type of approach leaves a lot to be desired with respect to portability. Another method which is attracting increasing attention is the chemometric or artificial intelligence system involving compensation by multi-sensor array configurations. Sensitivity and limit-of-detection have attracted little research due to the overwhelming nature of other difficulties. In the present paper we evaluate a number of these technical problems and discuss the architecture of devices that are currently available. Finally, some thoughts as to priorities for re-directing sensor research in the bioprobe area are presented.
OBJECTIVE: To program an artificial intelligence system, a neural network, and use it to predict results of sperm penetration in bovine cervical mucus (Penetrak assay; Serono Laboratories, Norwell, MA) and zona-free hamster egg penetration from the semen analysis. DESIGN: Results of 139 Penetrak assays, 1,416 zona-free hamster egg penetration assays, and the corresponding semen analyses were retrospectively analyzed by an artificial neural network. MAIN OUTCOME MEASURES: Classification errors of the neural network were compared with those of linear and quadratic discriminant function analyses. RESULTS: Data were separated into training and test sets. For the Penetrak result, linear and quadratic discriminant function analysis correctly predicted 58% and 74% of the training set results and only 64.1% and 69.2% of the test data, respectively. The neural network correctly predicted 92% of training set results and 80% of test set results. For the zona-free hamster egg penetration assay outcome, linear and quadratic discriminant function analysis correctly classified 66.3% and 46.0% of the training set and 64.9% and 44.7% of the test set, respectively. The neural network correctly classified 75.7% of the training data and 67.8% of the test data. CONCLUSIONS: Using the semen analysis, the neural network correctly classified 67.8% of zona-free hamster egg penetration assay results and 80% of Penetrak results it had not encountered previously, suggesting that this method of data analysis may be successfully employed to predict fertility potential.
PURPOSE: New methods are required to improve the prediction of cancer progression as traditional statistical tests have limited accuracy. Accurate predictions would allow physicians to offer specific treatment according to individual patient risk. While predictive improvements are obtained using ANN, the hidden nature of these networks prevents insight and has hindered their widespread implementation. NFM is an alternate form of artificial intelligence using fuzzy logic (which is a multivalued logic which provides reasoning under uncertainty). By defuzzification the NFM rule base becomes transparent to overcome the black box nature of ANN. MATERIALS AND METHODS: Combinations of clinicopathological (tumor stage and grade, patient age, gender, and smoking status) and molecular (immunohistochemical expression of p53 and methylation status of 11 loci) data from 117 patients were used to develop and compare predictive models of tumor progression using NFM, ANN and LR. RESULTS: NFM (88% to 100% sensitivity, 97% to 100% specificity and 94% to 100% accuracy) predicted the presence and timing of cancer progression more accurately than ANN (81% to 87%, 95% to 100% and 89% to 90%, p = 0.002) and LR 3%, 61% to 72% and 47% to 53%, p = 0.00005). NFM was able to interrogate the clinicopathological and molecular data, and select the most important parameters (age, grade, stage, smoking, methylation) for progression prediction. CONCLUSIONS: Intelligent systems and molecular biomarkers improved the accuracy of cancer progression predictions. NFM appeared superior to ANN in terms of accuracy, sensitivity, specificity and transparency. The use of NFM in routine clinical practice warrants further validation.
Development of urodynamics particularly during the past decade are highlighted, as well as a number of issues to be faced in the near future: research into the etiopathogenesis of different conditions, finding more sensitive and specific diagnostic procedures to overcome the current ones, establishing more comprehensive indications for urodynamics examinations and, as a result of all the above achieving greater improvement of certain surgical procedures. Both the technique and interpretation of the different urodynamic examinations, as well as the more recent innovations, implementation issues and controversies are detailed to a highly up-to-date level. Neuromodulation and ambulatory urodynamics deserve thoughtful consideration in this paper. Finally, attention is given to the controversies and future challenges such as urodynamic research providing accurate diagnosis of lower urinary tract obstruction in women, establishing the indication of urodynamic studies in women with urinary exertional incontinence, outlining the indications of neuromodulation and ambulatory urodynamics, applicability of artificial intelligence systems, improvement of artificial sphincter materials, tissue growth for bladder enlargement and actual prevention of myelodysplasia.
The reinforcement learning scheme proposed in Halici (J. Biosystems 40 (1997) 83) for the random neural network (RNN) (Neural Computation 1 (1989) 502) is based on reward and performs well for stationary environments. However, when the environment is not stationary it suffers from getting stuck to the previously learned action and extinction is not possible. To overcome the problem, the reinforcement scheme is extended in Halici (Eur. J. Oper. Res., 126(2000) 288) by introducing a new weight update rule (E-rule) which takes into consideration the internal expectation of reinforcement. Although the E-rule is proposed for the RNN, it can be used for training learning automata or other intelligent systems based on reinforcement learning. This paper looks into the behavior of the learning scheme with internal expectation for the environments where the reinforcement is obtained after a sequence of cascaded decisions. The simulation results have shown that the RNN learns well and extinction is possible even for the cases with several decision steps and with hundreds of possible decision paths.
Reasoning is the result of the computations made by intelligent systems, for instance those in the brain. It is not an abstract concept because calculations performed by computations are very concrete transactions among the different central processing unit components. Entropy measurements are proposed here to disclose the plasticity of the cerebral processing associated with language comprehension in video game playing. It is also assumed that entropy may be evaluated from the correlation coefficients obtained for the game event-related activity calculated for the different electroencephalogram derivations in the 10/20 system. The brain mapping derived from these entropy measurements clearly demonstrates the reallocation of speech functions to right brain areas when the classic left language circuits are damaged during prenatal life.