Search

A chunking procedure [Bradley, P.S. and Mangasarian, O.L., 2000, Massive data discrimination via linear support vector machines. Optimization Methods and Software, 13, 1-10. Available online at: ftp://ftp.cs.wisc.edu/mathprog/tech-reports/98-05.ps] utilized in [Mangasarian, O.L. and Thompson, M.E., 2006, Massive data classification via unconstrained support vector machines. Journal of Optimization Theory and Applications, 131, 315-325. Available online at: ftp://ftp.cs.wisc.edu/pub/dmi/tech-reports/06-01.pdf] for linear classifiers is proposed here for nonlinear kernel classification of massive datasets. A highly accurate algorithm based on nonlinear support vector machines that utilize a linear programming formulation [Mangasarian, O.L., 2000, Generalized support vector machines. In: A. Smola, P. Bartlett, B. Scholkopf and D. Schuurmans (Eds) Advances in Large Margin Classifiers (Cambridge, MA: MIT Press), pp. 135-146. Available online at: ftp://ftp.cs.wisc.edu/math-prog/tech-reports/98-14.ps] is developed here as a completely unconstrained minimization problem [Mangasarian, O.L., 2005, Exact 1-Norm support vector machines via unconstrained convex differentiable minimization. Technical Report 05-03, Data Mining Institute, Computer Sciences Department, University of Wisconsin, Madison, Wisconsin. Available online at: ftp://ftp.cs.wisc.edu/pub/dmi/tech-reports/05-03.ps. Journal of Machine Learning Research, 7, 1517-1530, 2006.]. This approach together with chunking leads to a simple and accurate method for generating nonlinear classifiers for a 250,000-point dataset that typically exceeds machine capacity when standard linear programming methods such as CPLEX [ILOG, 2003, ILOG CPLEX 9.0 User's Manual, Incline Village, Nevada. Available online at: http://www.ilog.com/products/cplex/] are used. Because a 1-norm support vector machine underlies the proposed method, the approach together with a reduced support vector machine formulation [Lee, Y.-J. and Mangasarian, O.L., 2001, RSVM: reduced support vector machines. Proceedings of the First SIAM International Conference on Data Mining, Chicago, 5-7 April, CD-ROM. Available online at: ftp://ftp.cs.wisc.edu/pub/dmi/tech-reports/00-07.ps] minimizes the number of kernel functions utilized to generate a simplified nonlinear classifier.

Background Advising on ill-health retirement is an important role of most practising occupational physicians. In recent years, the eligibility criteria and process for gaining early retirement benefits have changed in many pension schemes in the UK. Aim To investigate the variation in rates of retirement due to ill-health in National Health Service (NHS) Trusts and Local Authorities and to update previously published guidance on ill-health retirement with specific reference to pension schemes with eligibility criteria that include permanence of incapacity due to ill-health. Methods Rates of retirement were calculated for 222 NHS Trusts and 132 Local Authorities with more than 1500 employees. Literature searches and consensus statements by the authors. Results Rates of retirement were widely distributed in the NHS Trusts and Local Authorities. The median rates of retirement were 2.11 (IQR 1.37-2.91)/1000 active members and 4.10 (IQR 3.01-6.10)/1000 employees, respectively (P < 0.001). Difficulties in the doctor-patient relationship and in ascertaining the true functional ability of some patients were identified. Conclusion There continues to be marked variation in rates of early retirement due to ill-health within and between organizations that warrants further investigation. The general and specific guidance that appears as an appendix in Supplementary data to this paper should help occupational physicians to make equitable recommendations when assessing applications for early retirement benefits and fitness to work.

SummaryTreatment of 25 maturity-onset diabetic patients with diet and fenfluramine resulted in improved mean glucose tolerance and diabetic clinic random blood sugar levels compared with a longitudinal control group of diabetic patients treated with diet alone. Improved mean glucose tolerance occurred in both obese and non-obese patients in the fenfluramine group. Multiple regression analysis suggested that only part of the improvement of glucose tolerance in the fenfluramine group resulted from the increased weight loss associated with this treatment. No significant effect of fenfluramine treatment on oral glucose tolerance test plasma insulin levels was observed.ResumeLe traitement de 25 diabetiques de la maturitye par un regime et la fenfluramine a occasione une amelioration de la tolerance moyenne au glucose et de la glycemie mesuree au hasard au cours des consultations, par rapport a un groupe temoin longitudinal de diabetiques traites par le regime seul. l'amelioration de la tolerance moyenne au glu...

Acute experiments were performed on open-chest male mongrel dogs that were anesthetized with thiopental sodium and respired. An average of 10 penetrations per side was used to explore the rostral-caudal organization of the nucleus ambiguus (NA) and the area ventrolateral to it. Atrial pacing was performed on some trials to clarify the occurrence of atrioventricular (AV) block. In addition, the vagi were sectioned in some animals to confirm the peripheral pathway of the observed effects. Finally, the twin-pulse technique was used to classify the neurons underlying the observed effects according to their refractory periods. Stimulation in the area of the NA and lateral to it produced bradycardia and bradyarrhythmias without altering blood pressure. Stimulation of either the left or right NA produced slowing of both atrial and ventricular rates. However, slowing of the atrial rate was greatest with right NA stimulation, whereas slowing of the ventricular rate was predominant with left-side stimulation. The presence of AV block was evident from dropped QRS complexes with and without pacing. Vagotomy demonstrated that these effects were conducted by ipsilateral but not contralateral nerves. Measurement of the refractory periods showed that the neurons underlying both atrial and ventricular slowing were characteristic of B-fibers.

Oral glucose-tolerance test (O.G.T.T.) plasma sugar and insulin levels were measured in 118 newly diagnosed maturity-onset diabetic patients before and after treatment with diet alone for periods of 2 and 6 months. The results of glucose-tolerance tests carried out during treatment could be predicted from the initial test and the weight reduction between the tests. This prediction was not improved by the addition of further variables, including age, obesity, and plasma-insulin levels during the first test. The change in O.C.T.T. plasma-insulin between the first and second tests was predicted by the result of the initial tests, the improvement of glucose tolerance between the two tests, and the degree of weight reduction. 95% of the group achieved some improvement of glucose tolerance after 2 months of dietary treatment, and 59% of the group achieved adequate diabetic control by this time. It is concluded that treatment with diet alone should be the first-line management for patients with newly diagnosed maturity-onset diabetes mellitus.

1. Blood pressure was measured by random zero sphygmomanometer in the morning and afternoon for 5 days after normal delivery in a group of 136 previously normotensive women. The number of women studied each day varied from 32 to 125. 2. The afternoon blood pressure was higher than the morning blood pressure (differences: 1.7 mmHg systolic, 2.6 mmHg diastolic; P < 0.05). 3. Both systolic and diastolic blood pressures rose for the first 4 days after delivery. The average rise over the whole period was about 6 mmHg systolic and 4 mmHg diastolic (P < 0.05). 4. A considerable number of previously normotensive women displayed elevations of blood pressure in the puerperium. Twelve per cent of all patients exceeded a diastolic blood pressure of 100 mmHg.

Recordings were made from units in or near nucleus ambiguus (NA) in the awake, drug-free dog while arterial pressure (AP) was increased and decreased nonpharmacologically by inflating occlusion cuffs that had been placed around the descending aorta and inferior vena cava, respectively. Units were recorded differentially through floating fine-wire electrodes that extended from vertically movable cannulas in a chronically implanted microdrive. Fifty-three percent of units recorded in or near NA were excited by increased AP, but were unresponsive to decreased AP. The remaining units were unresponsive to increased or decreased AP. About 25% of AP-sensitive units had activity patterns related to the cardiac cycle. AP-sensitive, cardiac cycle related units located in NA displayed several of the properties previously demonstrated for antidromically indentified preganglionic parasympathetic cardiac motoneurons involved in baroreflex control of heart rate and rhythm. Recording from such neurons in the awake, neurologically intact, drug-free dog permits the fullest and most natural expression of neuro-cardiologic interactions and will allow the investigation of behavioral influences on autonomic and cardiovascular function.

A recent paper suggested that smoking does not increase medical care expenditures and, therefore, that reduced smoking is unlikely to decrease it. This conclusion differs from that of most other studies in this area and the reasons for this disagreement are investigated. It is pointed out that medical care expenditure estimates yield different results depending on whether they are estimated on a per capita basis or in total for populations which change in size as smoking is reduced or eliminated. To avoid disagreement in future studies of this type, some specific proposals are made which should facilitate the comparison of results obtained from different studies.

A longitudal study was carried out to separate the effects of treatment with diet from those of treatment with glibenclamide and phenformin on the oral glucose tolerance test (OGTT) plasma sugar and insulin levels in 118 patients with maturity-onset diabetes mellitus. Both drugs improved glucose tolerance and random blood sugar levels more than diet alone. The improved glucose tolerance with both drugs was due to a combination of a reduced fasting plasma sugar level and an improved tolerance to the glucose load. Treatment with glibenclamide increased insulin secretion but phenformin had no significant effect. Mean body weight fell slightly during phenformin administration, whereas a small but insignificant rise in patients receiving glibenclamide.

To the Editor.— The recent dramatic increases in the quality and availability of medical services in many Western countries have led to attempts to identify areas where the rise in costs may be slowed down without diminishing substantially the quality of services. One area of possible government intervention is the use of tobacco products, and particularly of cigarettes. If the purpose of such intervention is to help to keep health care costs at a manageable level and to prevent other economic losses caused by smoking, it becomes necessary to assess all the economic side effects as well. However, estimates of the economic costs of cigarette smoking in the United States vary widely, 1-5 partly because no two studies appear to have taken the same items into account, and partly because of the many difficulties inherent in arriving at such estimates. For example, even when only the direct costs of health