Predictions of coronary artery stenosis by artificial neural network.

Data from angiography patient records comprised 14 input variables of a neural network. Outcomes (coronary artery stenosis or none) formed both supervisory and output variables. The network was trained by backpropagation on 332 records, optimized on 331 subsequent records, and tested on final 100 records. If 0.40 was chosen as the output distinguishing stenosis from no stenosis, 81 patients who had stenosis would have been identified, while 9 of 19 patients who did not have stenosis might have been spared angiography. The results demonstrated that artificial neural networks could identify some patients who do not need coronary angiography.

Artificial neural network predictions of lengths of stay on a post-coronary care unit.

OBJECTIVE: To create and validate a model that predicts length of hospital unit stay. DESIGN: Ex post facto. Seventy-four independent admission variables in 15 general categories were utilized to predict possible stays of 1 to 20 days. SETTING: Laboratory. SAMPLE: Records of patients discharged from a post-coronary care unit in early 1993. RESULTS: An artificial neural network was trained on 629 records and tested on an additional 127 records of patients. The absolute disparity between the actual lengths of stays in the test records and the predictions of the network averaged 1.4 days per record, and the actual length of stay was predicted within 1 day 72% of the time. CONCLUSIONS: The artificial neural network demonstrated the capacity to utilize common patient admission characteristics to predict lengths of stay. This technology shows promise in aiding timely initiation of treatment and effective resource planning and cost control.

Myofibrillar and membrane-bound enzymes in skeletal muscle from myodystrophic mice.

1. Experiments were carried out to examine the biochemical changes, such as contractile protein biochemistry and membrane bound enzyme alterations associated with skeletal muscles of myd/myd. 2. Our studies demonstrate that there was a progressive decline in myofibrillar ATPase activity, and this decrease is greatest in 30 weeks old animals of myd/myd as compared to controls. 3. The proteolytic activity of myofibrils isolated from myd/myd was significantly higher than controls. 4. There was no significant difference in Ca2+ ATPase activity of myosin and actin-activated myosin ATPase activity of myd/myd and their controls. 5. Mg2+ ATPase and Na(+)+K(+)-ATPase of myodystrophic SL showed significant increase compared to controls. 6. Isoproterenol stimulated adenylate cyclase activity was significantly lower in the SL of dystrophic mice compared to controls. 7. GTP+isoproterenol stimulate adenylate cyclase was significantly higher in control SL and SR when compared to SL and SR isolated from myd/myd. 8. Guanylate cyclase activity was greater in myodystrophic mice both in the absence and presence of Triton X-100. cGMP and cAMP phosphodiesterase activities were greater in dystrophic mice as compared to controls. 9. These observations suggest that there are significant changes in myofibrillar ATPase, myofibrillar protease and membrane bound enzymes of myd/myd compared to control.

Additional force during stretch of single frog muscle fibres following tetanus.

Single fibres from the anterior tibialis muscles of frogs were used in paired difference experiments to investigate the long-term effect (up to 3.8 s) of tetanic stimulation on fibre stiffness during relaxation. The fibres were stretched from sarcomere lengths of 2.5 microns to 3.0 microns at constant velocities for periods ranging from 0.5 to 1.75 s. The first stretch of each pair took place when the fibres had not experienced a tetanus for at least 5 min. The second stretch took place 20-30 s after the first, but it was preceded by a tetanus (100 Hz stimulation applied for 190 ms; temperature, 23 degrees C). The force produced by the first stretch was subtracted from the force produced by the second to produce a paired difference. The fibres were held at the sarcomere length of 3.0 microns except for a brief period of time immediately prior to the stretches (1200 or 1450 ms). During those periods the fibres were shortened to 2.5 microns (250 ms) and then held at 2.5 microns, regardless of whether a tetanus was elicited, for either 950 or 1200 ms. The second stretch of each pair began either at the end of, or 250 ms after, the last stimulating pulse of the tetanus. At every velocity of stretch, the force produced by the fibres during the stretches was greater when the stretches were preceded by a tetanus than when they were not, and the additional force peaked at the conclusion of the stretches. The additional force, which was produced during the stretches following the tetani, declined for the remainder of the data acquisition period (up to 3 s) following the completion of stretches; it extrapolated to zero at 7-8 s after the completion of the stretches. The magnitude of the additional force was a non-linear direct function of the rate of stretch. Thixotropy or an increased stiffness to stretch was observed in all of the fibres following periods of quiescence.

Effects of organophosphorus agents on sarcoplasmic reticulum in skinned skeletal muscle fibers.

These experiments were designed to determine whether skinned skeletal muscle fibers could be useful in screening new antidotes to organophosphorus poisons. Isometric force and fiber diameter were measured in mechanically skinned fibers from mice and frogs. Fibers were depleted of calcium and placed in a calcium loading solution that contained 0.5 mM EGTA with pCa 6.25. The elapsed time (zero time) before a contracture began and the maximum rate of force development (slope) were measured and divided by the square of the diameter (normalized zero time, normalized slope). The zero time was assumed to be the time required for the sarcoplasmic reticulum to attain a threshold concentration for calcium-induced calcium release, and the slope was assumed to indicate primarily the rapidity of the release of calcium from the sarcoplasmic reticulum. Organophosphorus agents, sarin, soman, tabun, and VX were also placed in the loading solutions. Only sarin failed to shorten the normalized zero times of mouse fibers compared to controls, and all agents decreased the normalized slopes. The normalized zero times of frog fibers were not altered by the agents, but the normalized slopes were altered by some agents. Pralidoxime chloride (PAM) and 3-Cl-2,5,6-trimethylbenzoic acid (TBA) were also added to the loading solution for mouse fibers; PAM was marginally effective in moderating some actions of the organophosphates. Because the effects of the agents on the fibers were so definite, we concluded that the skinned muscle fiber might indeed be useful as a screening tool for developing and testing new antidotes to organophosphorus poisons.

Control of myofibrillar ATPase activity and force in myodystrophic muscle.

Myofibrillar ATPase activity was measured as a function of the free calcium concentration in skeletal muscles of control and myodystrophic mice. In addition, the force developed in skinned extensor digitorum longus (EDL) fibers of control and myodystrophic mice was measured as a function of the free calcium concentration, and a histomorphometric study was performed on soleus and EDL muscles of control and myodystrophic mice. The results showed that the myofibrillar ATPase activity and the force-generating mechanisms of control and myodystrophic muscles were controlled to the same relative degree by equivalent concentrations of calcium ions. Upon maximal activation of the ATPase activities, we measured 18% less activity in myodystrophic muscles than in control muscles. Maximal activation of the force-generating capacity in skinned fibers showed there was no significant difference in force produced in the control compared to myodystrophic fibers. The histomorphometric study revealed no alteration in the relative distribution of different fiber types in myodystrophic compared to control muscles. However, the histomorphometry did reveal a larger slow (type 1) relative cellular area compared to total cross-sectional area in myodystrophic muscle than in controls. We propose that the lower ATPase activity but equal force-generating capacity of myodystrophic muscles compared to control muscles is due to myodystrophic muscles being composed of a greater fraction of myofibrils from slow (type 1) fibers than control muscles.

Ca2+ capacity and uptake rate in skinned fibers of myodystrophic muscle.

Parameters related to the capacity and the rate of uptake of calcium ions by the sarcoplasmic reticulum were measured in skinned extensor digitorum longus fibers of control and myodystrophic mice. Single fibers were isolated and skinned in a relaxing solution and mounted on a force transducer and apparatus for changing the bathing solution (T = 25 degrees C). To test the capacity of the sarcoplasmic reticulum, fibers were placed in a solution for maximal loading and then moved to a test solution in which the major anion in the relaxing solution, gluconate, was replaced by chloride. In the resulting contractures, the means of the forces produced by 10 control and myodystrophic fibers were not significantly different. The conclusion is that the capacities of sarcoplasmic reticulum for calcium in control and myodystrophic fibers are equivalent. To test the rate of loading of sarcoplasmic reticulum, 11 control and myodystrophic fibers were depleted of calcium with caffeine and EGTA. Then they were placed in a solution with pCa = 5.5, and the delay before a contracture began was recorded. The delay was the time required for the sarcoplasmic reticulum to load calcium and attain a threshold for calcium-induced calcium release. The mean delay for the control fibers was significantly less than the mean delay in myodystrophic for the control fibers was significantly less than the mean delay in myodystrophic fibers. The disparity of loading times probably reflected a difference in the activities of the calcium pumps or a difference in the number of pump sites; 5 microM valinomycin did not significantly alter the loading times of either type or fiber.

Transverse impedance of single frog skeletal muscle fibers.

The transverse electrical impedance of single frog skeletal muscle fibers was measured at 31 frequencies that ranged from 1 to 100,000 Hz. Each fiber was bathed entirely in Ringer's solution, but it was positioned so that a central length of 5 mm was in a hollow plastic disk and was electrically isolated from the ends of the fiber. The diameter of the segment of the fiber in the disk was measured and then the segment was pressed from opposite sides by two insulating wedges. Electrical current was passed transversely through the segment between two platinum-platinum black electrodes that were located in the pools of Ringer's solution within the disk. The results were corrected for stray parallel capacitance, series resistance of the Ringer's solution between the fiber and the electrodes, parallel shunt resistance around the fiber, and the phase shift of the measuring apparatus. A nonlinear least-squares routine was used to fit a lumped equivalent circuit to the data from six fibers. The equivalent circuit that was chosen for the fibers contained three parallel branches; each branch was composed of a resistor and a capacitor in series. The model also included a seventh adjustable parameter that was designed to account for the degree of compression of the fibers by the insulating wedges. The branches of the equivalent circuit were assumed to represent the electrical properties of: (a) the myoplasm in series with the membrane capacitance that was exposed directly to the pools of Ringer's solution; (b) the capacitance and series resistance of the transverse tubules that were exposed directly to the pools of Ringer's solution; (c) the membrane capacitance in series with the shunt resistance between the fibers and the insulating wedges. The results gave no indication that current entered the sarcoplasmic reticulum.

Calcium ionophores and tension production in skinned frog muscle fibers.

Antibiotics, X-537A and A23187, were added in micromolar concentrations to selected bathing solutions of skinned frog muscle fibers, and they were shown to affect the production of tension in the skinned fibers. Segments of skinned fibers were bathed in a buffered calcium solution with a pCa near the threshold for contraction. When a segment was moved to a lightly buffered calcium solution with a pCa higher than the threshold for contraction but one containing either antibiotic, a transient contracture of the segment resulted. if the procedure was repeated in the same segment, no contracture was produced. The results are consistent with the idea that the antibiotics function as calcium ionophores in the membrane of the sarcoplasmic reticulum of skinned fibers.

Sizes of components in frog skeletal muscle measured by methods of stereology.

Stereological techniques of point and intersection counting were used to measure morphological parameters from light and electron micrographs of frog skeletal muscle. Results for sartorius muscle are as follows: myofibrils comprise 83% of fiber volume; their surface to volume ratio is 3.8 mum-1. Mitochondria comprise 1.6% of fiber volume. Transverse tubules comprise 0.32% of fiber volume, and their surface area per volume of fiber is 0.22 mum-1. Terminal cisternae of the sarcoplasmic reticulum comprise 4.1% of fiber volume; their surface area per volume of fiber is 0.54 mum-1. Longitudinal sarcoplasmic reticullum comprises 5.0% of fiber volume, and its surface area per volume of fiber is 1.48 mum-1. Longitudinal bridges between terminal cisternae on either side of a Z disk were observed infrequently; they make up only 0.035% of fiber volume and their surface area per volume of fiber is 0.009 mum-1. T-SR junction occurs over 67% of the surface of transverse tubules and over 27% of the surface of terminal cisternae. The surface to volume ratio of the caveolae is 48 mum-1; caveolae may increase the sarcolemmal surface area by 47%. Essentially the same results were obtained from semitendinosus fibers.

Longitudinal impedance of single frog muscle fibers.

The longitudinal impedance of single skeletal muscle fibers has been measured from1 to 10,000 Hz in an oil gap apparatus which forces current to flow longitudinally down the fiber. The impedance observed is purely resistive in some fibers from the semitendinosus muscle and in two fibers from the sartorius muscle. In other fibers from the semitendinosus muscle a small phase shift is observed. The mean value of the maximum phase shift observed from all fibers is 1.07 degrees. The artifacts associated with the apparatus and method are examined theoretically and it is shown that one of the likely artifacts could account for the small phase observed. It is concluded that the longitudinal impedance of skeletal muscle fibers is essentially resistive and that little, if any, longitudinal current crosses the membranes of the sarcoplasmic reticulum.

Size changes in single muscle fibers during fixation and embedding.

During fixation of single muscles fibers with glutaraldehyde, the volume of the fiber shrinks 20%, recovers in rinse and osmium tetroxide to near normal volume and shrinks 20% again when staining with uranyl acetate. This suggest that osmotic properties of membranes may not have been completely lost during fixation, post-fixation and en bloc staining. Dehydration in ethanol and propylene oxide produces a further 10% shrinkage in volume. Infiltration and embedding with Epon causes an additional 15% change in volume. This gives a total shrinkage in volume of 45% which is nearly twice that of the apparent shrinkage in the volume of the myosin lattice as determined by electron microscopy.

Longitudinal impedance of skinned frog muscle fibers.

Longitudinal impedance of skinned muscle fibers was measured with extracellular electrodes and an oil gap method in which a central longitudinal section of fiber is insulated by oil while the ends of the fiber are bathed in conducting pools of relaxing solution. Intact single fibers were isolated from frog semitendinosus muscle and the sarcolemma removed either by mechanical or chemical methods. Stray capacitance across the oil gap was measured after each experiment and its admittance subtracted from the admittance of the fiber and oil gap. Effects of impedance at the ends of the fiber were eliminated by measuring the impedance with two lengths of fiber in the oil gap and subtracting the impedance at the shorter length from that at the longer length. Longitudinal impedance so determined for mechanically and chemically skinned fibers exhibited zero phase shift from 1 to 10,000 Hz, i.e., the longitudinal impedance of skinned fibers is purely resistive. If we assume that our skinned fibers are a model of the sarcoplasm of muscle, we conclude that the equivalent circuit of the sarcoplasm is a resistor.

The surface area of sheep cardiac Purkinje fibres.

1. Measurements combining the techniques of point counting and line integration were performed on light and electron micrographs of Purkinje fibres from the sheep's heart. The measurements were aimed at determining membrane areas of importance for the cellular electrophysiology of this tissue.2. The mean volume fractions of the cells occupied by various constituents were: myofibrils, 0.234; mitochondria, 0.103; and nuclei, 0.009. The mean volume fraction of the fibres occupied by the interspaces between the tightly packed cells was 0.0023.3. The mean fractions of intercellular surface area occupied by junctional specializations were: nexus, 0.17; desmosome, 0.023; and fascia adherens, 0.014.4. The mean surface to volume ratio of the Purkinje cells and fibres was 0.46 mu(-1) which is 11.5 times the value of the surface to volume ratio of a long right circular cylinder 100 mu in diameter.5. There are two reasons for the increment in the surface to volume ratio of the fibre (when compared to that of a long right circular cylinder 100 mu in diameter): the multicellular composition of the fibres and the extensive folding of the surface of the cells.6. After correction for the intercellular nexal area the surface to volume ratio of a long cylindrical fibre 100 mu in diameter was 0.39 mu(-1), or about 10 times the value for a long right circular cylinder 100 mu in diameter. The surface to volume ratio of the tissue interspaces in the same fibre was 170 mu(-1).7. It was concluded that the total sarcolemmal area in this tissue is great enough so that the specific membrane capacitance could be about 1 muF/cm(2) and the specific membrane resistance 20,000 Omega cm(2).

The effect of potassium and chloride ions on the volume and membrane potential of single barnacle muscle cells.

1. In single barnacle skeletal muscle cells cell diameter has been measured as a function of external osmolality, and cell diameter and membrane potential have been measured during changes of external K and Cl concentrations ([K](o) and [Cl](o)) like those described in frog muscle by Hodgkin & Horowicz (1959). The diameter was monitored microscopically with a precision of 0.2-0.4% (S.D.). [K](o) was varied from 1 to 18 mM, a range of concentrations which does not cause contracture.2. At pH 8.0 the Cl permeability was so low that net KCl and water movements were absent. Such net movements were present at pH 4.5, corresponding to a change in the ratio (Cl conductance/K conductance) from approximately 1/12 at pH 8.0 to 1/2 at pH 4.5.3. Characteristically long time constants were observed for membrane potential responses to a change in [K](o) and/or [Cl](o), even at constant [K](o).[Cl](o). This phenomenon is attributed to a delayed equilibration by diffusion within the system of sarcolemmal invaginations and T-tubules. The delay in the response was increased by introducing polyvinylpyrrolidone (PVP) into this system, presumably because PVP raises intratubular viscosity.4. At pH 4.5 anomalous rectification for net movements of K was demonstrated by measurements of cell diameter and of membrane potential.

Magnesium in single skeletal muscle cells of Balanus.

Single skeletal muscle cells of Balanus contain 48 +/- 1 mmoles magnesium/kg dry weight. Although (28)Mg can be shown either to enter the cells or to be bound to the cell surface within less than 10 min, only 2.1 +/- 0.3% of cellular or cell surface Mg exchanges with this isotope even after several hours. Glycerinated cells washed out in Tris buffer at low ionic strength retain approximately 70% of the Mg present in intact cells. About 85% of this Mg is removed by extraction with KCl or NaCl at concentrations of K and Na which prevail in intact cells, as well as by pyrophosphate, Tris-ATP, or reduction of the ionized Mg concentration to 1 microM. Lowering the ionized Mg concentration to 0.1 microM does not further reduce the Mg content of glycerinated cells. The pH dependence of KCl-inextractable Mg suggests that more than one class of binding sites is involved. A significant fraction of the KCl-inextractable Mg bound to glycerinated cells fails to exchange with (28)Mg even after long equilibration. It is suggested that this fraction may be actin-bound Mg incorporated into the thin filaments during the polymerization of actin.