Application of multilayer feedforward neural networks toprecipitation cell-top altitude estimation

The use of passive 118-GHz O₂ observations of rain cells for precipitation cell-top altitude estimation is demonstrated by using a multilayer feedforward neural network retrieval system. Rain cell observations at 118 GHz were compared with estimates of the cell-top altitude obtained by optical stereoscopy. The observations were made with 2-4-km horizontal spatial resolution by using the millimeter-wave temperature sounder (MTS) scanning spectrometer aboard the NASA ER-2 research aircraft during the Genesis of Atlantic Lows Experiment (GALE) and the Cooperative Huntsville Meteorological Experiment (COHMEX) in 1986. The neural network estimator applied to MTS spectral differences between clouds, and nearby clear air yielded an rms discrepancy of 1.76 km for a combined cumulus, mature, and dissipating cell set and 1.44 km for the cumulus-only set. An improvement in rms discrepancy to 1.36 km was achieved by including additional MTS information on the absolute atmospheric temperature profile. An incremental method for training neural networks was developed that yielded robust results, despite the use of as few as 56 training spectra. Comparison of these results with a nonlinear statistical estimator shows that superior results can be obtained with a neural network retrieval system. Imagery of estimated cell-top altitudes was created from 118-GHz spectral imagery gathered from CAMEX, September through October 1993, and from cyclone Oliver, February 7, 1993     

38. Generalversammlung von CENELEC

Ohne Zusammenfassung     

On solving constrained optimization problems with neural networks:a penalty method approach

Deals with the use of neural networks to solve linear and nonlinear programming problems. The dynamics of these networks are analyzed. In particular, the dynamics of the canonical nonlinear programming circuit are analyzed. The circuit is shown to be a gradient system that seeks to minimize an unconstrained energy function that can be viewed as a penalty method approximation of the original problem. Next, the implementations that correspond to the dynamical canonical nonlinear programming circuit are examined. It is shown that the energy function that the system seeks to minimize is different than that of the canonical circuit, due to the saturation limits of op-amps in the circuit. It is also noted that this difference can cause the circuit to converge to a different state than the dynamical canonical circuit. To remedy this problem, a new circuit implementation is proposed.     

A case of macrocystic cervical neurinoma diagnosed by MRI with Gd-DTPA

The authors report a rare case of a large cystic cervical neurinoma. A 45-year-old female was admitted to our clinic because of motor weakness of the right upper extremity, numbness of the right fingers and right posterior cervical pain. Metrizamide CT myelography demonstrated the outline of a low density mass. MRI showed a mass revealing low signal intensity on T1-weighted image, high signal intensity on T2-weighted image and marginal enhancement on contrast image with Gd-DTPA. The mass which was diagnosed as cystic tumor, was located in the intradural extramedullary space between C4 to C5 segments. After C4 through C5 laminectomy, the tumor was found to originate from the C5 anterior motor root. The tumor consisted mostly of a cystic part with a very thin solid compartment beneath the capsule. Postoperative course of the patient was uneventful. Although spinal neurinoma is one of the most common spinal tumors, an almost completely degenerated large cystic spinal neurinoma is extremely rare. MRI with Gd-DTPA was useful for the diagnosis of the cystic neurinoma by clearly enhancing the margin of the tumor.     

Photometric application of the Gram stain method to characterize natural bacterial populations in aquatic environments

The Gram stain method was applied to the photometric characterization of aquatic bacterial populations with a charge-coupled device camera and an image analyzer. Escherichia coli and Bacillus subtilis were used as standards of typical gram-negative and gram-positive bacteria, respectively. A mounting agent to obtain clear images of Gram-stained bacteria on Nuclepore membrane filters was developed. The bacterial stainability by the Gram stain was indicated by the Gram stain index (GSI), which was applicable not only to the dichotomous classification of bacteria but also to the characterization of cell wall structure. The GSI spectra of natural bacterial populations in water with various levels of eutrophication showed a distinct profile, suggesting possible staining specificity that indicates the presence of a particular bacterial population in the aquatic environment.     

Influence of gas composition on recurrence of atelectasis after a reexpansion maneuver during general anesthesia

BACKGROUND: Atelectasis, an important cause of impaired gas exchange during general anesthesia, may be eliminated by a vital capacity maneuver. However, it is not clear whether such a maneuver will have a sustained effect. The aim of this study was to determine the impact of gas composition on reappearance of atelectasis and impairment of gas exchange after a vital capacity maneuver. METHODS: A consecutive sample of 12 adults with healthy lungs who were scheduled for elective surgery were studied. Thirty minutes after induction of anesthesia with fentanyl and propofol, the lungs were hyperinflated manually up to an airway pressure of 40 cmH2O. FIO2 was either kept at 0.4 (group 1, n = 6) or changed to 1.0 (group 2, n = 6) during the recruitment maneuver. Atelectasis was assessed by computed tomography. The amount of dense areas was measured at end-expiration in a transverse plane at the base of the lungs. The ventilation-perfusion distributions (VA/Q) were estimated with the multiple inert gas elimination technique. The static compliance of the total respiratory system (Crs) was measured with the flow interruption technique. RESULTS: In group 1 (FIO2 = 0.4), the recruitment maneuver virtually eliminated atelectasis for at least 40 min, reduced shunt (VA/Q < 0.005), and increased at the same time the relative perfusion to poorly ventilated lung units (0.005 < VA/Q < 0.1; mean values are given). The arterial oxygen tension (PaO2) increased from 137 mmHg (18.3 kPa) to 163 mmHg (21.7 kPa; before and 40 min after recruitment, respectively; P = 0.028). In contrast to these findings, atelectasis recurred within 5 min after recruitment in group 2 (FIO2 = 1.0). Comparing the values before and 40 min after recruitment, all parameters of VA/Q were unchanged. In both groups, Crs increased from 57.1/55.0 ml.cmH2O-1 (group 1/group 2) before to 70.1/67.4 ml.cmH2O-1 after the recruitment maneuver. Crs showed a slow decrease thereafter (40 min after recruitment: 61.4/60.0 ml.cmH2O-1), with no difference between the two groups. CONCLUSIONS: The composition of inspiratory gas plays an important role in the recurrence of collapse of previously reexpanded atelectatic lung tissue during general anesthesia in patients with healthy lungs. The reason for the instability of these lung units remains to be established. The change in the amount of atelectasis and shunt appears to be independent of the change in the compliance of the respiratory system.     

Structure and gene expression of avian cyclin D2

Avian cyclin D2 (Cyl D2)-encoding cDNA clones were isolated from a chicken UG9 T-cell lambda gt10 library. Sequence analysis revealed a high degree of sequence conservation with both the mouse and human Cyl D2, and somewhat lower similarity with the mouse and human Cyl D1 and D3. The homology is highest between species in the Cyl-box domain which is well conserved among human, mouse and chicken. A single 6.0-kb CYL2 mRNA is produced in both avian B- and T-cells, as expected.