Systematic tuning of parameters in support vector clustering

Clustering algorithms divide a set of observations into groups so that members of the same group share common features. In most of the algorithms, tunable parameters are set arbitrarily or by trial and error, resulting in less than optimal clustering. This paper presents a global optimization strategy for the systematic and optimal selection of parameter values associated with a clustering method. In the process, a performance criterion for the optimization model is proposed and benchmarked against popular performance criteria from the literature (namely, the Silhouette coefficient, Dunn's index, and Davies-Bouldin index). The tuning strategy is illustrated using the support vector clustering (SVC) algorithm and simulated annealing. In order to reduce the computational burden, the paper also proposes an alternative to the adjacency matrix method (used for the assignment of cluster labels), namely the contour plotting approach. Datasets tested include the iris and the thyroid datasets from the UCI repository, as well as lymphoma and breast cancer data. The optimal tuning parameters are determined efficiently, while the contour plotting approach leads to significant reductions in computational effort (CPU time) especially for large datasets. The performance criteria comparisons indicate mixed results. Specifically, the Silhouette coefficient and the Davies-Bouldin index perform better, while the Dunn's index is worse on average than the proposed performance index.     

Effects of lung volume on lung and chest wall mechanics in rats

To investigate the effect of lung volume onchest wall and lung mechanics in the rats, we measured theimpedance (Z) under closed- and open-chest conditions at variouspositive end-expiratory pressures (0-0.9 kPa) by using acomputer-controlled small-animal ventilator (T. F. Schuessler andJ. H. T. Bates. IEEE Trans. Biomed. Eng. 42: 860-866, 1995) that we have developed fordetermining accurately the respiratory Z in small animals. The Z oftotal respiratory system and lungs was measured with small-volumeoscillations between 0.25 and 9.125 Hz. The measured Z was fitted to amodel that featured a constant-phase tissue compartment (withdissipation and elastance characterized by constantsG andH, respectively) and a constant airwayresistance (Z. Hantos, B. Daroczy, B. Suki, S. Nagy, and J. J. Fredberg. J. Appl.Physiol. 72: 168-178, 1992). We matched the lungvolume between the closed- and open-chest conditions by using thequasi-static pressure-volume relationship of the lungs to calculate Zas a function of lung volume. Resistance decreased with lung volume andwas not significantly different between total respiratory system andlungs. However, G andH of the respiratory system weresignificantly higher than those of the lungs. We conclude that chestwall in rats has a significant influence on tissue mechanics of thetotal respiratory system.

     

A novel validation algorithm allows for automated cell tracking and the extraction of biologically meaningful parameters

Automated microscopy is currently the only method to non-invasively and label-free observe complex multi-cellular processes, such as cell migration, cell cycle, and cell differentiation. Extracting biological information from a time-series of micrographs requires each cell to be recognized and followed through sequential microscopic snapshots. Although recent attempts to automatize this process resulted in ever improving cell detection rates, manual identification of identical cells is still the most reliable technique. However, its tedious and subjective nature prevented tracking from becoming a standardized tool for the investigation of cell cultures. Here, we present a novel method to accomplish automated cell tracking with a reliability comparable to manual tracking. Previously, automated cell tracking could not rival the reliability of manual tracking because, in contrast to the human way of solving this task, none of the algorithms had an independent quality control mechanism; they missed validation. Thus, instead of trying to improve the cell detection or tracking rates, we proceeded from the idea to automatically inspect the tracking results and accept only those of high trustworthiness, while rejecting all other results. This validation algorithm works independently of the quality of cell detection and tracking through a systematic search for tracking errors. It is based only on very general assumptions about the spatiotemporal contiguity of cell paths. While traditional tracking often aims to yield genealogic information about single cells, the natural outcome of a validated cell tracking algorithm turns out to be a set of complete, but often unconnected cell paths, i.e. records of cells from mitosis to mitosis. This is a consequence of the fact that the validation algorithm takes complete paths as the unit of rejection/acceptance. The resulting set of complete paths can be used to automatically extract important biological parameters with high reliability and statistical significance. These include the distribution of life/cycle times and cell areas, as well as of the symmetry of cell divisions and motion analyses. The new algorithm thus allows for the quantification and parameterization of cell culture with unprecedented accuracy. To evaluate our validation algorithm, two large reference data sets were manually created. These data sets comprise more than 320,000 unstained adult pancreatic stem cells from rat, including 2592 mitotic events. The reference data sets specify every cell position and shape, and assign each cell to the correct branch of its genealogic tree. We provide these reference data sets for free use by others as a benchmark for the future improvement of automated tracking methods.     

Effect of gravity and posture on lung mechanics.

The volume-pressure relationship of the lung was studied in six subjects on changing the gravity vector during parabolic flights and body posture. Lung recoil pressure decreased by approximately 2.7 cmH(2)O going from 1 to 0 vertical acceleration (G(z)), whereas it increased by approximately 3.5 cmH(2)O in 30 degrees tilted head-up and supine postures. No substantial change was found going from 1 to 1.8 G(z). Matching the changes in volume-pressure relationships of the lung and chest wall (previous data), results in a decrease in functional respiratory capacity of approximately 580 ml at 0 G(z) relative to 1 G(z) and of approximately 1,200 ml going to supine posture. Microgravity causes a decrease in lung and chest wall recoil pressures as it removes most of the distortion of lung parenchyma and thorax induced by changing gravity field and/or posture. Hypergravity does not greatly affect respiratory mechanics, suggesting that mechanical distortion is close to maximum already at 1 G(z). The end-expiratory volume during quiet breathing corresponds to the mechanical functional residual capacity in each condition.     

Effects of stent design parameters on normal artery wall mechanics

A stent is a device designed to restore flow through constricted arteries. These tubular scaffold devices are delivered to the afflicted region and deployed using minimally invasive techniques. Stents must have sufficient radial strength to prop the diseased artery open. The presence of a stent can subject the artery to abnormally high stresses that can trigger adverse biologic responses culminating in restenosis. The primary aim of this investigation was to investigate the effects of varying stent "design parameters" on the stress field induced in the normal artery wall and the radial displacement achieved by the stent. The generic stent models were designed to represent a sample of the attributes incorporated in present commercially available stents. Each stent was deployed in a homogeneous, nonlinear hyperelastic artery model and evaluated using commercially available finite element analysis software. Of the designs investigated herein, those employing large axial strut spacing, blunted corners, and higher amplitudes in the ring segments induced high circumferential stresses over smaller areas of the artery's inner surface than all other configurations. Axial strut spacing was the dominant parameter in this study, i.e., all designs employing a small stent strut spacing induced higher stresses over larger areas than designs employing the large strut spacing. Increasing either radius of curvature or strut amplitude generally resulted in smaller areas exposed to high stresses. At larger strut spacing, sensitivity to radius of curvature was increased in comparison to the small strut spacing. With the larger strut spacing designs, the effects of varying amplitude could be offset by varying the radius of curvature and vice versa. The range of minimum radial displacements from the unstented diastolic radius observed among all designs was less than 90 microm. Evidence presented herein suggests that stent designs incorporating large axial strut spacing, blunted corners at bends, and higher amplitudes exposed smaller regions of the artery to high stresses, while maintaining a radial displacement that should be sufficient to restore adequate flow.     

Genetic and molecular coordinates of neuroendocrine lung tumors,with emphasis on small-cell lung carcinomas

The aim of this review is to present the advances in our understanding of the progression of tumorigenesis in neuroendocrine lung tumors. Current information on established and putative diagnostic and prognostic markers of neuroendocrine tumors are evaluated, with a special reference to small-cell lung carcinoma, due to its higher incidence and aggressive behavior. The genetic and molecular changes that accompany these neoplasms are highlighted, and factors that influence cell-cycle progression, apoptosis, drug resistance, and escape from immune surveillance are critically assessed.     

Frictional contact mechanics methods for soft materials: application to tracking breast cancers

Mammography is currently the most widely used screening and diagnostic tool for breast cancer. Because X-ray images are 2D projections of a 3D object, it is not trivial to localise features identified in mammogram pairs within the breast volume. Furthermore, mammograms represent highly deformed configurations of the breast under compression, thus the tumour localisation process relies on the clinician's experience. Biomechanical models of the breast undergoing mammographic compressions have been developed to overcome this limitation. In this study, we present the development of a modelling framework that implements Coulomb's frictional law with a finite element analysis using a C(1)-continuous Hermite mesh. We compared two methods of this contact mechanics implementation: the penalty method, and the augmented Lagrangian method, the latter of which is more accurate but computationally more expensive compared to the former. Simulation results were compared with experimental data from a soft silicon gel phantom in order to evaluate the modelling accuracy of each method. Both methods resulted in surface-deformation root-mean-square errors of less than 2mm, whilst the maximum internal marker prediction error was less than 3mm when simulating two mammographic-like compressions. Simulation results were confirmed using the augmented Lagrangian method, which provided similar accuracy. We conclude that contact mechanics on soft elastic materials using the penalty method with an appropriate choice of the penalty parameters provides sufficient accuracy (with contact constraints suitably enforced), and may thus be useful for tracking breast tumours between clinical images.     

Effects of hypoxia on lung mechanics in the newborn cat

The present study was designed to investigate the effects of hypoxia on lung mechanics in the newborn cat and to determine if vagal efferent innervation to the airways is involved in the response. We studied 11 animals, aged 2-7 days, anesthetized with a mixture of chloralose-urethane administered intraperitoneally. A tracheal cannula was inserted just below the larynx and following paralysis (pancuronium bromide), mechanical ventilation was initiated. A pneumothorax was created by a midline thoracotomy and an end-expiratory load was applied to maintain functional residual capacity. Animals were placed in a flow plethysmograph from which measurements of transpulmonary pressure, flow, and volume, mean inspiratory resistance, and dynamic compliance of the lung were calculated. The experimental protocol consisted of a series of 8-min trials, each preceded by a controlled volume history. The hypoxia challenge was composed of 1 min of ventilation with 40% O2, followed by 5 min exposure to 10% O2 and 2 min of recovery. In the majority of animals (7 out of 11), hypoxia had no effect on lung mechanics compared with control trials. Four animals responded to hypoxia with an increase in resistance and a decrease in compliance. Resistance remained elevated throughout the hypoxia with an average maximal increase of 47.2 +/- 22.2% (SD). Dynamic compliance was significantly decreased at the 2nd, 3rd, and 4th min only of hypoxia. Bilateral vagotomy abolished the response in the four animals and hypoxia had no effect on mechanics postvagotomy. Our data suggest that in most cases changes in lung mechanics do not play a causal role in the biphasic ventilatory response to hypoxia seen in the newborn.