Panel discussion: how to monitor and minimize variation and mistakes.
Explore the source record for details and available documents.
Centers for Disease Control and Prevention, U.S.↗
Biomedical subjects
M L Astion
Publications and source records attributed to M L Astion.
Evaluation of an interactive computer program to teach gram-stain interpretation.
Explore the source record for details and available documents.
Computer-Assisted Instruction↗
A computer program that periodically monitors the ability to interpret the antinuclear antibody test.
Our laboratory has been developing computer programs that help medical technologists improve their performance of the microscope-based immunofluorescence assay for antinuclear antibodies (ANA). This image-based laboratory test has been associated with poor reproducibility. We have previously described our first program, ANA-Tutor, which systematically teaches the ANA test by using approximately 150 processed digital images of ANA test results. The program we describe here, Pattern Plus Auditor, is a logical extension to ANA-Tutor. Pattern Plus Auditor tests the ability of laboratory personnel to interpret the ANA test, and tracks individual and laboratory performance over time. The program consists of image-based questions that test a variety of ANA staining patterns, including homogeneous, speckled, centromere, nucleolar, mixed patterns, and rare patterns. For each question, the program provides correct answers with explanations and color overlays that highlight key image features. By entering the proper password, users gain access to exam results for individuals and for the laboratory as a whole. Results are available for the current exam, any previous exam, or cumulatively on all exams to date. Intralaboratory testing with computer programs such as Pattern Plus Auditor might be a useful part of quality-assurance procedures for many image-based laboratory tests.
Antibodies, Antinuclear↗
Electrophoresis-tutor: an image-based personal computer program that teaches clinical interpretation of protein electrophoresis patterns of serum, urine, and cerebrospinal fluid.
High-resolution protein electrophoresis of serum, urine, and cerebrospinal fluid (CSF) can aid in the diagnosis of multiple myeloma, amyloidosis, macroglobulinemia, multiple sclerosis, and other diseases. Electrophoresis-Tutor is a personal computer program based on approximately 150 digital images that teaches the clinical interpretation of agarose gel electrophoretic patterns. The program is divided into the following sections: introduction, CSF, serum, urine, review of disease states, program navigator, and final exam. The CSF section describes normal and abnormal CSF findings with emphasis on oligoclonal banding, as seen in the CSF of patients with multiple sclerosis. The serum section emphasizes monoclonal gammopathy patterns but also has detailed descriptions of inflammation, liver disease, protein-losing disorders, genetic deficiencies, and other patterns. Monoclonal gammopathy is described in the context of specific associated clinical conditions (e.g., myeloma, amyloidosis). For each monoclonal gammopathy example, results of standard electrophoresis, densitometry, and immunofixation are presented. The review of disease states uses animation to illustrate the development and remission of a variety of pathological patterns. The program navigator allows the user to jump quickly to any place in the program. The optional exam contains 20 questions, and detailed feedback is given after each question. Electrophoresis-Tutor can be used as a stand-alone teaching tool, a companion to traditional instruction, or a reference source.
Blood Proteins↗
BibleCard: network-based virtual database for laboratory information.
The clinical laboratory's use of computers has evolved beyond the single minicomputer stand-alone system. Our laboratory information system is now part of an institutional network. The laboratory also uses smaller systems and workstations for a wide variety of functions, often with much data duplication among systems. We have been developing a network-based virtual database for laboratory test information. This system uses World Wide Web standards for hypertext and multimedia displays, which allows for the display of information retrieved from various department computer sources without the necessity of data duplication, modification of existing systems, or centralization of data. The medical technologists can continue to write testing procedures on their word processors. Maintenance of reference values, specimen requirements, etc., can continue as a laboratory information system function. Yet information from all of these disparate sources can be viewed in a consolidated format that has platform independence.
Chemistry, Clinical↗
Application of neural networks to the classification of giant cell arteritis.
OBJECTIVE: Neural networks are a group of computer-based pattern recognition methods that have recently been applied to clinical diagnosis and classification. In this study, we applied one type of neural network, the backpropagation network, to the diagnostic classification of giant cell arteritis (GCA). METHODS: The analysis was performed on the 807 cases in the vasculitis database of the American College of Rheumatology. Classification was based on the 8 clinical criteria previously used for classification of this data set: 1) age > or = 50 years, 2) new localized headache, 3) temporal artery tenderness or decrease in temporal artery pulse, 4) polymyalgia rheumatica, 5) abnormal result on artery biopsy, 6) erythrocyte sedimentation rate > or = 50 mm/hour, 7) scalp tenderness or nodules, and 8) claudication of the jaw, of the tongue, or on swallowing. To avoid overtraining, network training was terminated when the generalization error reached a minimum. True cross-validation classification rates were obtained. RESULTS: Neural networks correctly classified 94.4% of the GCA cases (n = 214) and 91.9% of the other vasculitis cases (n = 593). In comparison, classification trees correctly classified 91.6% of the GCA cases and 93.4% of the other vasculitis cases. Neural nets and classification trees were compared by receiver operating characteristic (ROC) analysis. The ROC curves for the two methods crossed, indicating that the better classification method depended on the choice of decision threshold. At a decision threshold that gave equal costs to percentage increases in false-positive and false-negative results, the methods were not significantly different in their performance (P = 0.45). CONCLUSION: Neural networks are a potentially useful method for developing diagnostic classification rules from clinical data.
Diagnosis, Computer-Assisted↗
Neural networks as expert systems in rheumatic disease diagnosis: artificial intelligence or intelligent artifice?
Explore the source record for details and available documents.
Artificial Intelligence↗
Overtraining in neural networks that interpret clinical data.
Backpropagation neural networks are a computer-based pattern-recognition method that has been applied to the interpretation of clinical data. Unlike rule-based pattern recognition, backpropagation networks learn by being repetitively trained with examples of the patterns to be differentiated. We describe and analyze the phenomenon of overtraining in backpropagation networks. Overtraining refers to the reduction in generalization ability that can occur as networks are trained. The clinical application we used was the differentiation of giant cell arteritis (GCA) from other forms of vasculitis (OTH) based on results for 807 patients (593 OTH, 214 GCA) and eight clinical predictor variables. The 807 cases were randomly assigned to either a training set with 404 cases or to a cross-validation set with the remaining 403 cases. The cross-validation set was used to monitor generalization during training. Results were obtained for eight networks, each derived from a different random assignment of the 807 cases. Training error monotonically decreased during training. In contrast, the cross-validation error usually reached a minimum early in training while the training error was still decreasing. Training beyond the minimum cross-validation error was associated with an increased cross-validation error. The shape of the cross-validation error curve and the point during training corresponding to the minimum cross-validation error varied with the composition of the data sets and the training conditions. The study indicates that training error is not a reliable indicator of a network's ability to generalize. To find the point during training when a network generalizes best, one must monitor cross-validation error separately.
Bias↗
The application of backpropagation neural networks to problems in pathology and laboratory medicine.
Neural networks are a group of computer-based pattern recognition technologies that have been applied to problems in clinical diagnosis. This review focuses on one member of the group of neural networks, the backpropagation network. The steps in creating a backpropagation network are (1) collecting adequate training facts, (2) choosing the specific network structure, (3) training the network, and (4) cross-validating the trained network. The first published applications of backpropagation networks to problems in pathology and laboratory medicine have appeared recently. These applications are in the areas of image analysis and interpretation of laboratory results, and they demonstrate the feasibility of the approach.
Evaluation Studies as Topic↗
Application of neural networks to the interpretation of laboratory data in cancer diagnosis.
Neural networks are a relatively new method of multivariate analysis. The purpose of this study was to investigate the ability of neural networks to differentiate benign from malignant breast conditions on the basis of the pattern of nine variables: patient age, total cholesterol, high-density lipoprotein cholesterol, triglycerides, apolipoprotein A-I, apolipoprotein B, albumin, the tumor marker CA15-3, and the Fossel index (measurement of methylene and methyl line-widths in proton NMR spectra). The laboratory analyses were made with blood plasma or serum specimens. The neural network was "trained" with 57 patients: 23 patients with breast malignancies and 34 patients with benign breast conditions. A neural network with nine input neurons, 15 hidden neurons, and two output neurons correctly classified all 57 patients. The ability of the network to predict the diagnoses of patients that it had no encountered in training was tested with a separate group (cross-validation group) of 20 patients. The network correctly predicted the diagnoses for 80% of these patients. For comparison we analyzed the same sets of 57 training patients and 20 cross-validation patients by quadratic discriminant function analysis. The quadratic discriminant function, calculated from the same 57 patients used to train the neural network, correctly classified 84% of the 57 patients, and correctly diagnosed 75% of the 20 cross-validation patients. The results suggest that neural networks are a potentially useful multivariate method for optimizing the diagnostic utility of laboratory data.
Antigens, Tumor-Associated, Carbohydrate↗
Localization and stoichiometry of electrogenic sodium bicarbonate cotransport in retinal glial cells.
An electrogenic Na+/HCO3- cotransport system was identified and characterized in freshly dissociated salamander Müller (glial) cells. Under voltage-clamp, these cells generated an outward current when external HCO3- concentration [( HCO3-]o) was raised. This current was Na(+)-dependent, Cl(-)-independent, and was blocked by the stilbenes 4,4'-diisothiocyanato-stilbene-2,2'-disulfonate (DIDS) and 4,4'-dinitrostilbene-2,2'-disulfonate (DNDS), and by harmaline, demonstrating that the current was generated by a Na+/HCO3- cotransport system. Substantially larger currents were evoked when [HCO3-]o was raised at the Müller cell endfoot as compared to other cell regions, indicating that cotransporter sites are localized preferentially to the endfoot. The reversal potential of the current, which varied as a function of HCO3- and Na+ transmembrane gradients, indicated that the cotransporter has a HCO3-:Na+ stoichiometry of 3:1.
4,4'-Diisothiocyanostilbene-2,2'-Disulfonic Acid↗
Further studies of electrogenic Na+/HCO3- cotransport in glial cells of Necturus optic nerve: regulation of pHi.
In the presence of Ba++, an increase in the bath HCO3- at constant CO2 (i.e., variable bath pH) produced a hyperpolarization. The hyperpolarizing effect of adding HCO3-/CO2 at constant bath pH was not significantly affected by the presence of 50 mumol/l strophanthidin. In the absence of Ba++, addition of HCO3-/CO2 at constant bath pH produced a Na(+)-dependent hyperpolarization. Therefore, CO2 movements, electrogenic Na+/K+ pump activity and changes in Ba++ binding do not contribute significantly to the hyperpolarization induced by HCO3-. These results along with the results of previous studies (Astion et al: J Gen Physiol 93:731, 1989) strongly suggest that the hyperpolarization induced by the addition of HCO3- is due to an electrogenic Na+/HCO3- cotransporter, which transports Na+, HCO3- (or its equivalent), and net negative charge across the glial membrane. To study the role of electrogenic Na+/HCO3- cotransport in the regulation of pHi in glial cells, we used intracellular double-barreled, pH-sensitive microelectrodes. At a bath pH of 7.5, the mean initial intracellular pH (pHi) was 7.32 (SD 0.03, n = 6) in HEPES-buffered Ringer's solution and 7.39 (SD 0.1, n = 6) in HCO3-/CO2 buffered solution. These values for pHi are more than 1.2 pH units alkaline to the pHi predicted from a passive distribution of protons; thus, these cells actively regulate pHi. Superfusion and withdrawal of 15 mmol/l NH4+ induced an acidification of 0.2 to 0.3 pH units, which recovered toward the original steady-state pHi.(ABSTRACT TRUNCATED AT 250 WORDS)
Animals↗
Na+/H+ exchange in glial cells of Necturus optic nerve.
Single and double-barreled pH-sensitive electrodes were used to study intracellular pH (pHi) regulation in glial cells of Necturus optic nerve in the nominal absence of HCO3-/CO2. After the cells were acidified by the addition and withdrawal of NH4+, the pHi recovered toward the original steady-state pHi. The recovery from acidification was Na+-dependent and inhibited by 1 mM amiloride. These results suggest the existence in intact vertebrate glial cells of a Na+/H+ exchanger which functions in acid extrusion.
Animals↗
Facilitation of voltage-gated ion channels in frog neuroglia by nerve impulses.
The functions of glial cells in the nervous system are not well defined, with the exception of myelin production by oligodendrocytes, uptake of amino-acid synaptic transmitters, and a contribution to extracellular potassium homeostasis. Neuroglia have receptors for neurotransmitters which may be involved in neuron-glia interactions. Recent studies have demonstrated voltage-gated ion channels in glial membranes. In a study of the optic nerve of the frog, small areas of the surface were examined with the loose patch-clamp method, and voltage-gated Na+ and K+ channels, presumably located in the membranes of the astrocytes forming the glia limitans, were identified. We now report that nerve impulses in the axons of the frog optic nerve transiently alter the properties of the voltage-dependent membrane channels of the surface glial cells (astrocytes), a demonstration of a new form of neuron-glia interaction.
Action Potentials↗
Effects of barium and bicarbonate on glial cells of Necturus optic nerve. Studies with microelectrodes and voltage-sensitive dyes.
We have studied the effects of Ba++, a known K+ channel blocker, on the electrophysiological properties of the glial cells of Necturus optic nerve. The addition of Ba++ reversibly depolarized glial cells by 25-50 mV; the half maximal deplorization was obtained with a Ba++ concentration of approximately 0.3 mM. In the presence of Ba++, the sensitivity of the membrane to changes in K+ was reduced and there was evidence of competition between K+ and Ba++ for the K+ channel. These effects, which were accompanied by a large increase in the input resistance of the glial cells, indicate that Ba++ blocks the K+ conductance in glial cells of Necturus optic nerve. With the K+ conductance reduced, we were able to investigate the presence of other membrane conductances. We found that in the presence of Ba++, the addition of HCO3- caused a Na+-dependent hyperpolarization that was sensitive to the disulfonic stilbene SITS (4-acetamido-4'-isothiocyanostilbene-2, 2'-disulfonic acid). Removal of Na+ resulted in a HCO3- -dependent, SITS-sensitive depolarization. These results are consistent with the presence in the glial membrane of an electrogenic Na+/HCO3- cotransporter in which Na+, HCO3-, and net negative charge are transported in the same direction. In Cl- -free solutions, the Ba++-induced depolarization increased, suggesting a small permeability to Cl-. Using voltage-sensitive dyes and a photodiode array for multiple site optical recording, the distribution of potential changes in response to square pulses of intracellularly injected current were recorded before and after the addition of increased and the decay of amplitude as a function of distance decreased. Such results indicate that Ba++ increases the membrane resistance more than the resistance of the intercellular junctions.
Animals↗
Electrogenic Na+/HCO3- cotransport in neuroglia.
Membrane potential recording from glial cells in Necturus optic nerve in the presence of 2 mM Ba++, which was added to block the K+ conductance, gave the following results. 1) In HCO3- -free, low-Na+ solutions (11% of control; Na+ replaced with N-methyl-D-glucamine), the hyperpolarizing effect of adding 10 mM HCO3- was reduced by approximately 80%. 2) 4-acetamido-4'-isothiocyanato-stilbene-2,2'-disulfonic acid (SITS, 0.1 or 1 mM) reduced the effect of HCO3- by at least 50%. 3) In the presence of HCO3-, reduction of Na+ caused a depolarization which was much larger than that observed in nominally HCO3- -free solutions. These observations indicate the presence in the glial membrane of an electrogenic Na+/HCO3- cotransporter in which the stoichiometry of HCO3- to Na+ is greater than 1.
Animals↗
K+ accumulation in the space between giant axon and Schwann cell in the squid Alloteuthis. Effects of changes in osmolarity.
In a train of impulses in squid giant axon, accumulation of extracellular potassium causes successive afterhyperpolarizations to be progressively less negative. In Loligo, Frankenhaeuser and Hodgkin had satisfactorily accounted for the characteristics of this effect with a model in which the axon is surrounded by a space, width theta, and a barrier of permeability P. In axons isolated from Alloteuthis, we found that the model fitted the observations quite well. Superfusing the axon with hypotonic artificial seawater (ASW) caused theta and P to decrease, and, conversely, hypertonic ASW caused them to increase: this would be the case if both the space and the pathway through the barrier were extracellular. In some cases, in normal ASW, the afterhyperpolarizations in a train decreased very little, less than 0.7 mV. In these extreme cases, theta was estimated to be 190 nm and P to be 7 x 10(-4) cm s-1, both several times the values of 30 nm and 6 x 10(-5) cm s-1 estimated by Frankenhaeuser and Hodgkin. We suggest that in vivo the periaxonal space may be considerably wider than that seen in conventionally fixed squid tissue.
Action Potentials↗
Effects of bicarbonate on glial cell membrane potential in Necturus optic nerve.
Intracellular electrodes were used to continuously monitor the membrane potential of glial cells in the isolated Necturus optic nerve. Addition of up to 10 mM extracellular bicarbonate (with CO2), at constant pH, produced a hyperpolarization of up to 10 mV (with a time course almost as fast as that of a K+ depolarization) that returned toward baseline during the following 2-15 min. Upon bicarbonate withdrawal, the potential transiently became more positive. The bicarbonate effects were magnified when the K+ conductance was decreased and the cell depolarized by the addition of barium. Similar bicarbonate effects were observed in Cl- free solutions. These results suggest to us that: glial cells have a bicarbonate permeability of the same order as that to K+ and glial cells buffer transient changes in acid base balance in the neuronal microenvironment at the expense of their internal pH.
Animals↗