Robust deep learning object recognition models rely on low frequency information in natural images

Machine learning models have difficulty generalizing to data outside of the distribution they were trained on. In particular, vision models are usually vulnerable to adversarial attacks or common corruptions, to which the human visual system is robust. Recent studies have found that regularizing machine learning models to favor brain-like representations can improve model robustness, but it is unclear why. We hypothesize that the increased model robustness is partly due to the low spatial frequency preference inherited from the neural representation. We tested this simple hypothesis with several frequency-oriented analyses, including the design and use of hybrid images to probe model frequency sensitivity directly. We also examined many other publicly available robust models that were trained on adversarial images or with data augmentation, and found that all these robust models showed a greater preference to low spatial frequency information. We show that preprocessing by blurring can serve as a defense mechanism against both adversarial attacks and common corruptions, further confirming our hypothesis and demonstrating the utility of low spatial frequency information in robust object recognition.

Five points to check when comparing visual perception in humans and machines

With the rise of machines to human-level performance in complex recognition tasks, a growing amount of work is directed toward comparing information processing in humans and machines. These studies are an exciting chance to learn about one system by studying the other. Here, we propose ideas on how to design, conduct, and interpret experiments such that they adequately support the investigation of mechanisms when comparing human and machine perception. We demonstrate and apply these ideas through three case studies. The first case study shows how human bias can affect the interpretation of results and that several analytic tools can help to overcome this human reference point. In the second case study, we highlight the difference between necessary and sufficient mechanisms in visual reasoning tasks. Thereby, we show that contrary to previous suggestions, feedback mechanisms might not be necessary for the tasks in question. The third case study highlights the importance of aligning experimental conditions. We find that a previously observed difference in object recognition does not hold when adapting the experiment to make conditions more equitable between humans and machines. In presenting a checklist for comparative studies of visual reasoning in humans and machines, we hope to highlight how to overcome potential pitfalls in design and inference.

Learning to represent signals spike by spike

Networks based on coordinated spike coding can encode information with high efficiency in the spike trains of individual neurons. These networks exhibit single-neuron variability and tuning curves as typically observed in cortex, but paradoxically coincide with a precise, non-redundant spike-based population code. However, it has remained unclear whether the specific synaptic connectivities required in these networks can be learnt with local learning rules. Here, we show how to learn the required architecture. Using coding efficiency as an objective, we derive spike-timing-dependent learning rules for a recurrent neural network, and we provide exact solutions for the networks’ convergence to an optimal state. As a result, we deduce an entire network from its input distribution and a firing cost. After learning, basic biophysical quantities such as voltages, firing thresholds, excitation, inhibition, or spikes acquire precise functional interpretations.

Spiking neural networks can encode information with high efficiency in the spike trains of individual neurons if the synaptic weights between neurons are set to specific, optimal values. In this regime, the networks exhibit irregular spike trains, high trial-to-trial variability, and stimulus tuning, as typically observed in cortex. The strong variability on the level of single neurons paradoxically coincides with a precise, non-redundant, and spike-based population code. However, it has remained unclear whether the specific synaptic connectivities required in these spiking networks can be learnt with local learning rules. In this study, we show how the required architecture can be learnt. We derive local and biophysically plausible learning rules for recurrent neural networks from first principles. We show both mathematically and using numerical simulations that these learning rules drive the networks into the optimal state, and we show that the optimal state is governed by the statistics of the input signals. After learning, the voltages of individual neurons can be interpreted as measuring the instantaneous error of the code, given by the error between the desired output signal and the actual output signal.

Biomechanical Texture Coding in Rat Whiskers

Classically, texture discrimination has been thought to be based on ‘global’ codes, i.e. frequency (signal analysis based on Fourier analysis) or intensity (signal analysis based on averaging), which both rely on integration of the vibrotactile signal across time and/or space. Recently, a novel ‘local’ coding scheme based on the waveform of frictional movements, discrete short lasting kinematic events (i.e. stick-slip movements called slips) has been formulated. We performed biomechanical measurements of relative movements of a rat vibrissa across sandpapers of different roughness. We find that the classic global codes convey some information about texture identity, but are consistently outperformed by the slip-based local code. Moreover, the slip code also surpasses the global ones in coding for active scanning parameters. This is remarkable as it suggests that the slip code would explicitly allow the whisking rat to optimize perception by selecting goal-specific scanning strategies.

Demixed principal component analysis of neural population data

Neurons in higher cortical areas, such as the prefrontal cortex, are often tuned to a variety of sensory and motor variables, and are therefore said to display mixed selectivity. This complexity of single neuron responses can obscure what information these areas represent and how it is represented. Here we demonstrate the advantages of a new dimensionality reduction technique, demixed principal component analysis (dPCA), that decomposes population activity into a few components. In addition to systematically capturing the majority of the variance of the data, dPCA also exposes the dependence of the neural representation on task parameters such as stimuli, decisions, or rewards. To illustrate our method we reanalyze population data from four datasets comprising different species, different cortical areas and different experimental tasks. In each case, dPCA provides a concise way of visualizing the data that summarizes the task-dependent features of the population response in a single figure.

DOI:http://dx.doi.org/10.7554/eLife.10989.001

Many neuroscience experiments today involve using electrodes to record from the brain of an animal, such as a mouse or a monkey, while the animal performs a task. The goal of such experiments is to understand how a particular brain region works. However, modern experimental techniques allow the activity of hundreds of neurons to be recorded simultaneously. Analysing such large amounts of data then becomes a challenge in itself.

This is particularly true for brain regions such as the prefrontal cortex that are involved in the cognitive processes that allow an animal to acquire knowledge. Individual neurons in the prefrontal cortex encode many different types of information relevant to a given task. Imagine, for example, that an animal has to select one of two objects to obtain a reward. The same group of prefrontal cortex neurons will encode the object presented to the animal, the animal’s decision and its confidence in that decision. This simultaneous representation of different elements of a task is called a ‘mixed’ representation, and is difficult to analyse.

Kobak, Brendel et al. have now developed a data analysis tool that can ‘demix’ neural activity. The tool breaks down the activity of a population of neurons into its individual components. Each of these relates to only a single aspect of the task and is thus easier to interpret. Information about stimuli, for example, is distinguished from information about the animal’s confidence levels.

Kobak, Brendel et al. used the demixing tool to reanalyse existing datasets recorded from several different animals, tasks and brain regions. In each case, the tool provided a complete, concise and transparent summary of the data. The next steps will be to apply the analysis tool to new datasets to see how well it performs in practice. At a technical level, the tool could also be extended in a number of different directions to enable it to deal with more complicated experimental designs in future.

DOI:http://dx.doi.org/10.7554/eLife.10989.002

Effect of shock waves and cisplatin on cisplatin-sensitive and -resistant rodent tumors in vivo

The effect of a combination of shock waves with cisplatin was examined in vivo with subcutaneously implanted amelanotic melanomas (A-Mel 3) in Syrian golden hamsters and cisplatin-sensitive or cisplatin-resistant fibrosarcoma (SSK2/0 and SSK2/R2) in C3H mice. In all 3 tumor models, 4 treatment modalities were compared: control, cisplatin treatment, shock waves and the combination of shock waves and cisplatin. Shock waves significantly delayed tumor growth in all 3 tumor models when compared to the respective control group. Cisplatin alone delayed the growth of A-Mel 3 and SSK2/0, whereas SSK2/R2 remained uninfluenced by the drug. In all 3 tumor models the combined treatment with shock waves and cisplatin additively and significantly delayed tumor growth. In A-Mel-3-bearing animals the combined treatment significantly increased survival time. The growth of SSK2/0 and SSK2/R2 tumors was delayed to a similar extent by the combined treatment modality as compared to shock-wave treatment alone. This indicates that the cisplatin resistance of SSK2/R2 tumors has been overcome by the simultaneous shock wave treatment. An increased intracellular cisplatin accumulation in the tumors due to shock wave exposure is suggested as the mechanism of interaction between shock waves and cisplatin.

[Enzyme liberation and activation of the kallikrein-kinin system in experimental pancreatitis. Studies of portal vein blood, pancreatic lymph and peritoneal effusion]

The clinical course of acute pancreatitis is strongly influenced by secondary cardiac, pulmonary and renal damage. The aim of the present study was to gather information about the compartment promoting the systemic damage. Therefore the activity of lipase, phospholipase A and plasma pro-kallikrein and the concentration of tissue kallikrein and kininogen were measured in portal venous blood, pancreatic lymph and peritoneal exudate. Anaesthetized pigs were subjected to fluid resuscitation to keep systemic haemodynamic parameters constant. The pancreas was isolated in situ. The pigs were randomly assigned to a control group (n = 9) or one of the two pancreatitis groups (n = 10 each). Pancreatitis was induced by i.a. infusion of free fatty acid (FFS) or retrograde infusion of 5% sodium taurocholate intraductally (NaT). In both pancreatitis groups the activity of lipase and phospholipase A increased. The most pronounced changes were seen in the peritoneal exudate (phospholipase A activity 40 min after induction: control 10.0 U/l, NaT 72.2 U/l). In both pancreatitis groups there was evidence for activation of the tissue kallikrein kinin system in the form of an increase in the kallikrein concentration and a decrease in the kininogen concentration. Again the changes were most pronounced in the peritoneal exudate (tissue kallikrein 40 min after induction: control 14.7 ng/ml, NaT 452 ng/ml).

Dilation of coronary microvessels by adenosine induced hypotension in dogs

An experimental model was established for fluorescence video microscopy of coronary microvessels. Nineteen dogs were anesthetized with a narcotic. Catheters were placed for hemodynamic monitoring and sampling of arterial and coronary venous blood. Myocardial perfusion was measured with radioactive microspheres. Following thoracotomy, movements of the myocardial surface area under investigation were restricted by a specially designed heart holder. Plasma was stained with FITC labelled dextran. Diameters were determined in arteriolar and venular microvessels greater than or equal to 20 microns. Measurements were performed during baseline conditions, i.e. only the basic anesthetic drug was applied, and during coronary vasodilation by continuous infusion of adenosine in a randomized sequence. Mean arterial pressure was reduced from 85 +/- 2 mmHg during baseline to 59 +/- 1 mmHg by infusion of 16.9 +/- 2.2 mg.kg-1.h-1 adenosine. Adenosine increased left ventricular blood flow by 253%, left ventricular oxygen demand remained unchanged. A total of 495 arteriolar and 170 venular diameters were measured during baseline condition and during adenosine infusion. Arteriolar diameters increased in all vessel segments between 20 and 600 microns, however, arterioles below a critical size of 100 microns had a greater dilating capacity than larger arterioles. Maximal decrease of segmental resistance occurred in 20-40 microns arterioles and amounted to 74%, which is less than the 82% decrease of total coronary resistance. Venular diameter changes, too, were more pronounced in smaller vessels.

Myocardial oxygenation and transmural lactate metabolism during experimental acute coronary stenosis in pigs

Measurement of surface tissue pO2 (ptO2) with surface electrodes is increasingly applied in experimental medicine. Its use on the beating heart may seem to be problematic because transmural gradients of tissue pO2 would reduce the validity of pO2 determinations in the epicardial layers. This study attempted to determine whether ptO2 may be a valid and sensitive indicator of transmural myocardial oxygenation. In order to measure ptO2, two eight-channel Clark-type electrodes were placed on a beating porcine left ventricle (n = 13). Measurements were made at different degrees of acute stenosis of the left anterior descending artery (LAD). A 24-F cannula was inserted into the great cardiac vein, draining the poststenotic myocardium to obtain coronary venous blood samples. Transmural metabolic changes were detected simultaneously by coronary venous blood gas parameters and lactate levels. Epicardial tissue pO2 was 49 +/- 2 mm Hg (mean +/- SEM) before stenosis and decreased to a mean value of 25 +/- 2 mm Hg during stenosis. Different degrees of LAD stenosis (ptO2 range: 12-35 mm Hg) were substantial enough to alter arterio-coronary venous lactate difference (avd lactate) from +0.31 +/- 0.07 mmol/l (control) to -0.62 +/- 0.15 mmol/l (stenosis). A significant linear correlation between changes of ptO2 (delta ptO2) and changes of avd lactate (delta avd lactate) resulted (y = 0.59 + 0.62x; r = 0.86; p less than or equal to 0.001). However, linear regression analysis between delta ptO2 correlated with the corresponding data from coronary venous pO2 (delta pO2cv) oxygen content (delta O2contcv), and oxygen saturation (delta O2satcv) showed no significant correlations. We conclude that measurement of ptO2 is a sensitive and valuable indicator of transmural oxygenation in ischemic myocardium, whereas pO2cv, O2contcv and O2satcv do not seem to be valid predictors of ischemia in myocardial oxygenation.

Oleic acid induced pancreatitis in pigs

An experimental model of edematous pancreatitis in pigs was established and measurement of pancreatic macro- and microcirculatory parameters and determinations of pancreatic enzymes (lipase, phospholipase A) and vasoactive mediators (prostanoids, kallikrein, kininogen) were performed. During general anesthesia the pancreas was isolated in situ. Pancreatic microcirculatory parameters were measured using videofluorescence microscopy after iv administration of FITC-Dextran. In hourly collected samples lipase and phospholipase A activities were determined enzymatically, concentrations of kallikrein, kininogen, and selected prostanoids were measured by radioimmunoassay. Two experimental groups were studied: (1) control (n = 9); (2) edematous pancreatitis induced by injection of oleic acid into the pancreatic artery (free fatty acid, ffa; n = 10). The animals were followed up for 6 hr. Systemic hemodynamic parameters remained constant in both groups. In the pancreatitis group pancreatic blood flow and O2-consumption decreased significantly (-55 and -49%), while pancreatic vascular resistance increased significantly (+50%). During baseline conditions 41% of all capillaries were perfused. In the pancreatitis group there were both areas with persistent stasis as well as areas with continuous perfusion. However, in the latter areas the portion of perfused capillaries decreased significantly to 27%. In the control group the portion of perfused capillaries remained constant. Liberation of lipase and phospholipase A especially into lymph and ascites fluid was measured during pancreatitis. Furthermore, considerable releases of kallikrein into lymph (+50%) and ascites (+800%) and a marked consumption of kininogen in lymph (+90%) and in ascites fluid (+80%) were measured. Activation of the arachidonic acid cascade and a significant release of prostacyclin and thromboxane A2 into pancreatic venous blood and lymph was observed.(ABSTRACT TRUNCATED AT 250 WORDS)

Pulmonary hypertension after heparin-protamine: roles of left-sided infusion, histamine, and platelet-activating factor

Severe pulmonary hypertension after protamine neutralization of heparin is an infrequent but life-threatening event following cardiopulmonary bypass. The effect of left ventricular infusion of protamine on pulmonary hypertension as well as a possible role of platelet-activating factor (PAF) or histamine in the heparin-protamine reaction was investigated in 30 pigs in four different groups during general anesthesia. Group 1 animals received 250 IU/kg heparin, followed by 100 mg protamine intravenously after 15 min. In group 2 protamine was infused into the left ventricle. Group 3 animals received the histamine H1- and H2-antagonists clemastine and ranitidine 5 min before protamine infusion. In group 4 the PAF receptor blocker WEB 2086 was given 5 min before protamine. Platelet-activating factor was measured by a bioassay in serum samples of group 1 and group 4 animals. In all four groups protamine caused severe pulmonary hypertension, thromboxane A2 release, and a transient decrease in leukocyte counts. No PAF release was detected after protamine infusion. Neither left ventricular infusion of protamine nor histamine or PAF antagonists prevented or attenuated the reactions after protamine infusion. The authors conclude that left ventricular infusion of protamine provides no protection from pulmonary hypertension, and that histamine and PAF are not involved in the acute pulmonary vasoconstriction after protamine neutralization of heparin.

Proliferation rates of lymphocytes in subcutaneously transposed spleen and peripheral blood after heterotopic heart transplantation in rats

In a transplantation model using Dark Agouti rats as heart donors and Lewis rats as recipients, mean graft survival time was 7.1 days in GII, receiving no immunosuppression, and 31.4 days in GIII, animals immunosuppressed by DSG. A higher percentage of Lb in the spleen than in the PB in the transplanted groups was detected on certain days. HCT sharply decreased in immunosuppressed animals, thus suggesting a reversible suppression of erythropoesis induced by DSG.

Early gallstone recurrence rate after successful shock-wave therapy

Extracorporeal shock-wave lithotripsy combined with adjuvant bile-acid dissolution therapy results in complete clearance of stone fragments in a high percentage of selected patients with radiolucent gallbladder calculi. With the gallbladder in situ, these patients are at risk of stone recurrence. Therefore, the early rate of stone recurrence after successful lithotripsy was evaluated. Fifty-eight of the first 60 consecutive patients who became stone free underwent follow-up examinations at least 1 yr (range, 12-37 mo; mean +/- SD, 18 +/- 6) after discontinuation of adjuvant bile-acid therapy. Five patients reported recurrent biliary pain within 1 yr after lithotripsy, and recurrent gallstones were detected. Fifty-three patients were asymptomatic during the first yr, and no recurrence was detected. Thus, the rate of gallstone recurrence was 9% within 1 yr. The rate of gallstone recurrence up to 3 yr was estimated by actuarial analysis. The probability of stone recurrence was 11% (+/- 4%) at 1.5 yr, and no further increase was observed up to 3 yr. Gallstone recurrence within 1 yr after successful shock-wave therapy has to be expected in approximately the same percentage of patients as has been reported in earlier postdissolution trials. It causes recurrent biliary pain in most cases.

Epicardial oxygen tensions during changes in arterial PO2 in pigs

Arterial hypoxemia decreased epicardial tissue PO2, measured by means of a multiwire surface electrode, as well as coronary venous PO2 and myocardial lactate extraction. Left ventricular blood flow increased, O2 delivery, O2 demand and O2 consumption of the left ventricle remained unchanged. Thus, epicardial and coronary venous PO2 indicated decreased capillary and interstitial PO2 rather than cellular hypoxia. A linear relation between mean epicardial PO2 and coronary venous PO2 proves both parameters equally effective in reflecting changes in myocardial tissue oxygenation. However, PO2 distribution curves provide additional information and epicardial PO2 is superior in models with regional changes of myocardial oxygenation.

[Pancreatic circulation in experimental biliary pancreatitis]

Measurements of pancreatic micro- and macrocirculation were performed to evaluate the pancreatitis-induced changes. Pigs were anesthetized and ventilated mechanically. Hypotension induced side-effects were avoided by adequate volume replacement. After laparatomy, splenectomy and gastroectomy the animals were enterotomized. Systemic hemodynamic parameters were monitored as well as pancreatic blood flow (Q), which was measured electromagnetically, and arterial and portal-venous blood gases. Pancreatic microcirculatory parameters were observed using fluorescence-videomicroscopy after i.v. administration of FITC dextran 150 and FITC labeled autologous erythrocytes. The pigs were randomly assigned to a control (n = 9) and a pancreatitis group (n = 10), the later being induced by the retrograde infusion of sodium-taurocholate. Systemic and pancreatic macrohemodynamic parameters remained constant in both groups, except for avdO2 and O2-consumption (O2-c) decreasing significantly in the pancreatitis group. At baseline 42% of all capillaries were perfused in both groups. In pancreatitis we detected focal areas with persistent stasis and areas which were continuously perfused. In these areas the portion of capillaries perfused by erythrocytes increased significantly to 67%. This was accompanied by an extravasation of FITC dextran. The finding of an unchanged Q beside reduced O2-c and avdO2 during pancreatitis is explained by the changes in pancreatic microcirculation. Focal stasis was observed beside areas showing typical signs of an acute inflammation: increased macromolecular permeability and capillary recruitment, e.g. oedema and hyperaemia.

Release of arachidonic acid metabolites during acute pancreatitis in pigs

The pancreatic release of arachidonic acid metabolites was studied in a porcine model of acute pancreatitis. In situ isolation of the pancreatic gland enabled selective collection of pancreatic venous blood, pancreatic lymph, and ascites fluid. Three experimental groups were studied: 1) control (n = 9); 2) hemorrhagic pancreatitis induced by injection of 5% bile salt (sodium taurocholate) into the pancreatic duct (n = 10); and 3) edematous pancreatitis induced by injection of free fatty acid (FFA) into the pancreatic artery (n = 10). Determinations of cyclooxygenase metabolites were performed by radioimmunoassay; lipoxygenase metabolites (LTC4, LTD4) were measured by radioimmunoassay after purification by high-performance liquid chromatography. Prostaglandin (PG)F1 alpha, thromboxane B2, and PGF2 alpha concentrations were almost doubled in the lymph of the FFA group during pancreatitis, as were PGF1 alpha levels in pancreatic venous blood. However, concentrations of cyclooxygenase metabolites remained unchanged in the control group and in the bile salt group. Concentrations of LTC4 and LTD4 in lymph and ascites fluid of both pancreatitis groups increased from about 50 pg/ml to a mean level of 600 pg/ml at 6 h. Leukotriene concentrations in the control group were consistently below 50 pg/ml. The results of this study indicate that above all LTC4 and LTD4 are released from the organ and that these arachidonic acid metabolites may be also involved in the events following acute pancreatitis contributing to the systemic effects of the disease.

[Photodynamic laser tumor therapy with a hematoporphyrin derivative in spinocellular carcinoma of the external ear]

Haematoporphyrine-derivative (hpd) selectively photosensitises malignant tumours following intravenous application. 48 hours after injection of hpd an integral laser-light application of the tumour and the surrounding normal tissue was performed. We report on the successful administration of this photodynamic therapy in a patient with a spinocellular carcinoma of the auricle. 16 days after therapy the tumour showed a complete response without damage of the also irradiated surrounding normal tissue. Apart from a temporary sensitivity to light no side effects of the therapy were seen.

In vitro cytotoxic activity of lithotripter shock waves combined with adriamycin or with cisplatin on L1210 mouse leukemia cells

The effect of a combined treatment with shock waves generated by a lithotripter and Adriamycin or cisplatin was examined in cells that acutely survived exposure to shock waves and proliferated afterwards. Batches of 2 x 10(6) cells were exposed to the respective drug for 50 min or for 50 min plus 72 h. During the 50-min drug exposure 500 shock waves were applied at 25 kV. The growth as a percentage of the control was determined after 72 h by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. Cells treated with shock waves alone showed a growth inhibition as compared to control cells. For a 50-min drug exposure with Adriamycin the dose enhancement ratio did not exceed 1.3. For a 50-min drug exposure with cisplatin at concentrations of 0.5 micrograms/ml and 5.0 micrograms/ml, growth (as a percentage of the control) after combined treatment was significantly reduced as compared to cisplatin treatment alone; the dose enhancement ratio was 3.2 at 50% growth compared to the control. This indicates that shock waves can increase the susceptibility of L1210 cells to cisplatin. For a 50-min plus 72-h drug exposure no effect of an additional treatment with shock waves, as compared to chemotherapy alone, could be observed.

Myocardial surface PO2--an indicator of myocardial tissue oxygenation?

The validity of myocardial surface tissue PO2 (PtO2) as a reliable indicator of transmural myocardial tissue oxygenation was studied in six anaesthetised, open chest pigs. Epicardial surface PtO2 was correlated with other variables of myocardial tissue oxygenation such as regional blood flow, coronary venous PO2, O2 saturation, PCO2 and regional myocardial lactate extraction. The study design was based on an experimental model in which the effects of a pacing induced tachycardia on tissue oxygenation of ischaemic and normally supplied myocardium were measured. Two platinum multiwire surface electrodes were placed on the epicardium, on the areas supplied by the left anterior descending coronary artery (LAD) and the left circumflex coronary artery (CX). The LAD was constricted to reduce mean surface PtO2 in the LAD area to about 50% of its baseline value. This did not affect surface PtO2 in the CX area. The reduction of surface PtO2 in the LAD area was associated with decreases in coronary venous PO2 and O2 saturation and with increases in coronary venous lactate and PCO2. Subendocardial regional blood flow and the subendocardial to subepicardial flow ratio were significantly lower than in the CX area. Increasing the heart rate by pacing (+45 beats.min-1) led to an increased degree of ischaemia as shown by fall in surface PtO2 in the LAD area to values around zero kPa, by marked increase in coronary venous lactate and PCO2, by reduction in total (-10%) and subendocardial (-40%) LAD flow and by deterioration of the subendocardial to subepicardial flow ratio. The increased degree of ischaemia was not accompanied by an increase in O2 extraction. The marked decrease in surface PtO2 occurred in spite of a slight increase in the subepicardial regional blood flow (+10%); thus the increase in O2 delivery was not sufficient to meet the increase in O2 demand. Total flow was increased by 27% in the CX area without changes in the subendocardial to subepicardial flow ratio and in the surface PtO2 values. When pacing was stopped, surface values of PtO2 in the LAD area returned to prepacing values, as did lactate extraction and coronary venous PCO2. Clear and close relationships with surface PtO2 were found for regional lactate extraction, coronary venous PCO2 and the normalised subendocardial RBF. Poor or no correlations were found for the normalised subepicardial regional blood flow, the coronary venous O2 saturation and the absolute values of subendocardial and subepicardial regional blood flow.(ABSTRACT TRUNCATED AT 400 WORDS)

[Gd-DTPA-supported magnetic resonance tomographic perfusion follow-up of shockwave-treated tumors]

The signal characteristics of 14 shockwave-treated and 14 solid control tumors were studied before and after injection of Gd-DTPA in an animal model. T1-weighted images of shockwave-treated tumors documented no significant signal intensity increase after contrast media injection in comparison with the untreated control tumors. The reduction of perfusion in shockwave-treated tumors can be documented in vivo by the signal intensity changes of the tumors after contrast media injection.

Regional blood flow and tissue oxygen pressures of the collateral-dependent myocardium during isoflurane anesthesia in dogs

The authors investigated the effects of isoflurane on blood flow and tissue oxygen pressures of a collateral-dependent myocardium. Seventeen dogs divided into two groups were studied 3-4 weeks after implantation of ameroid coronary artery constrictors to completely occlude the proximal part of the left anterior descending artery. Experiments were performed during anesthesia with an opiate that was infused intravenously throughout the experiments. In Group 1 (n = 9), measurements were obtained during control and during isoflurane- (1.6-2.2 vol%) induced hypotension (mean arterial pressure, 60 mmHg). In Group 2 (n = 8), the identical protocol was applied, but norepinephrine was infused to maintain normotension. Dipyridamole effects were studied in five animals of Group 2 after a second control period at least 1 h after discontinuation of isoflurane. Isoflurane-induced hypotension caused reductions of blood flow and surface tissue oxygen pressures in the collateral flow-dependent area. Vasodilation in the normal left ventricular areas was demonstrated by an unchanged blood flow despite a reduced oxygen consumption and by a significantly increased coronary sinus hemoglobin oxygen saturation. When arterial pressure was maintained at its control level by norepinephrine, tissue oxygen pressures remained constant and collateral as well as normal area flow increased significantly during isoflurane. Coronary vascular resistance was lower during administration of isoflurane and norepinephrine compared with that during isoflurane induced hypotension, suggesting a significant contribution of tissue oxygen demand in regulation of coronary vascular resistance. At comparable levels of arterial pressure and left ventricular oxygen consumption, normal zone blood flow was significantly higher during dipyridamole than during isoflurane and norepinephrine. Thus, isoflurane-induced hypotension decreased blood flow and tissue oxygen pressures of collateral flow-dependent myocardial areas. However, neither isoflurane nor dipyridamole caused such alterations when arterial pressure was normal.

Left ventricular oxygen tensions in dogs during coronary vasodilation by enflurane, isoflurane and dipyridamole

The purpose of this study was to investigate the effects of the anesthetics enflurane and isoflurane and of the coronary vasodilator dipyridamole on myocardial oxygen balance and myocardial tissue oxygen tensions. The studies were performed in 24 open-chest dogs during basal anesthesia with a narcotic. Myocardial blood flow (MBF) was measured using radioactive microspheres, myocardial surface tissue PO2 by means of a platinum multiwire surface electrode. One control group and three experimental groups were studied: enflurane (1.1 vol%), isoflurane (0.7 vol%, both end-tidal concentrations), and dipyridamole (0.4 mg/kg). Mean arterial pressure significantly decreased to an average of 70 mm Hg in all three experimental groups. Although MBF was unchanged during enflurane (-18%) and isoflurane (+20%), it increased during dipyridamole (+304% p less than 0.05 vs baseline and control, enflurane, and isoflurane groups). Myocardial oxygen consumption decreased significantly during enflurane and isoflurane but remained unchanged during dipyridamole. Thus, the ratio between myocardial oxygen delivery and consumption increased 6% with enflurane (p less than 0.05 vs baseline), 47% with isoflurane (p less than 0.05 vs baseline and control group) and 280% with dipyridamole (p less than 0.05 vs baseline and control, enflurane, and isoflurane groups). Coronary venous PO2 remained unchanged during enflurane but increased significantly during isoflurane and dipyridamole. Left ventricular surface tissue PO2 was unchanged in enflurane and isoflurane animals and decreased slightly, yet significantly, during dipyridamole. All variables remained unchanged in the control group. Thus, isoflurane and dipyridamole interfered with MBF autoregulation and increased myocardial oxygen delivery out of proportion to myocardial demands.(ABSTRACT TRUNCATED AT 250 WORDS)

Myocardial contractility, blood flow, and oxygen consumption in healthy dogs during anesthesia with isoflurane or enflurane

Left ventricular contractility (Vmax), myocardial blood flow (MBF), and oxygen consumption (O2C) were determined together with systemic hemodynamic parameters in a total of 21 mongrel dogs. Baseline recordings were obtained under basal anesthetic conditions with a narcotic (piritramid, IV). In the control group (n = 7), recordings were obtained during a three-hour observation period with infusion of piritramid. In experimental groups measurements were repeated with equi-anesthetic concentrations of isoflurane (0.7 and 1.4 vol%; n = 8) and enflurane (1.1 and 2.2 vol%; n = 6). Dose-dependent reductions of arterial pressure, cardiac output (CO) and peripheral vascular resistance were observed with isoflurane and enflurane. CO at the higher anesthetic level was depressed significantly more with enflurane. This difference was obviously due to a more severe depression of myocardial contractility with enflurane; Vmax was decreased by 18% and 26% with enflurane, but only by 10% and 17% with isoflurane (P less than 0.01). MBF and the fraction of CO received by the heart were increased above their baseline values with both concentrations of isoflurane. In contrast, the fraction of CO remained constant with enflurane while MBF decreased. O2C was reduced due to decreases of afterload and left ventricular contractility. The reduction was greater with enflurane than with isoflurane. All parameters remained unchanged in the control group. The results of this study indicate that the most striking difference in the actions of isoflurane and enflurane on cardiac parameters is on myocardial vascular resistance; MBF is increased with isoflurane, but is decreased with enflurane although myocardial perfusion pressure is reduced by almost identical amounts.(ABSTRACT TRUNCATED AT 250 WORDS)

A new method of investigating the efficacy of regional thermotherapy in subcutaneous xenografts of nude mice

A new method of investigating the efficacy of regional thermotherapy in subcutaneous xenografts of nude mice is reported. The use of high frequency hyperthermia was well tolerated by the sensitive animals and allowed an exact continuous temperature measurement in different tumor regions. The interstitial procedure in this model could be the best approach for later clinical use in urology, e.g. for prostate treatment and is an alternative to the transrectal hyperthermia application.

Thromboxane mediation of pulmonary hemodynamic responses after neutralization of heparin by protamine in pigs

Protamine neutralization of heparin is often associated with severe hemodynamic side-effects, including pulmonary hypertension and systemic hypotension. Because prostanoids may be involved, the authors studied the role of arachidonic acid metabolites, especially thromboxane A2, in this process. During anesthesia with enflurane and fentanyl, four groups of pigs were studied: Group 1 (n = 10) received heparin (250 IU/kg), followed by protamine (100 mg) after 15 minutes to neutralize the heparin. The same protocol was used in group 2 (n = 11), except that the thromboxane A2 receptor antagonist BM 13.177 (10 mg/kg) was infused 5 minutes before the protamine. The protocol for group 1 was also used for group 3 (n = 7) except that these animals were pretreated with indomethacin (10 mg/kg). Animals in group 4 (n = 10) were given protamine only (100 mg). Pulmonary artery pressure and pulmonary vascular resistance increased significantly in group 1 after protamine neutralization of heparin. This was accompanied by significant increases in plasma concentrations of the cyclooxygenase products thromboxane B2, 6-keto-prostaglandin F1 alpha, and prostaglandin F2 alpha. Cyclooxygenase products increased to comparable degrees in group 2, but without hemodynamic effects. Leukocyte counts decreased comparably in both groups. Hemodynamic reactions, as well as changes in plasma prostanoid levels were absent in group 3, and group 4, but leukocyte counts were less affected in animals that received protamine alone. The results indicate that the hemodynamic side-effects of protamine are mediated by prostanoids and that thromboxane A2 release is the pivotal step, because side effects were effectively prevented by pretreatment with a thromboxane receptor antagonist.(ABSTRACT TRUNCATED AT 250 WORDS)

Splanchnic oxygen consumption and hepatic surface oxygen tensions during isoflurane anesthesia

Blood flow to and oxygen consumption of the splanchnic organs were determined together with hepatic surface oxygen tensions in 18 mongrel dogs anesthetized with the long-acting narcotic piritramid. Twelve animals also received 0.7 Vol% and 1.4 Vol% isoflurane; six time-related controls received piritramid only. Surgical preparation consisted of a left thoracotomy for inserting a catheter into the left atrium for microsphere injections and for gaining access to the hepatic surface through an incision in the diaphragm. Parameters in the animals receiving isoflurane were recorded at three stages: stage 1--piritramid anesthesia after surgical preparation; stage 2-60 min after addition of 0.7 Vol% (end-expiratory) isoflurane; stage 3-60 min after addition of 1.4 Vol% (end-expiratory) isoflurane. Hepatic surface oxygen tension was determined at each stage using an eight-channel oxygen sensitive electrode. Mean arterial pressure and cardiac output decreased during both stages with isoflurane; hepatic arterial inflow remained constant. Portal blood flow and, hence, total hepatic inflow decreased significantly. An unchanged splanchnic O2 consumption induced lower hepatic venous pO2 values: 40 +/- 1 mmHg at control, 35 +/- 2 mmHg, and 31 +/- 2 mmHg (mean +/- SEM; both P less than 0.05) during isoflurane. A concomitant decrease of hepatic surface pO2 values indicated an altered tissue oxygenation. The percentage of hepatic surface pO2 values in the lowest pO2 range (0-5 mmHg) increased significantly from 8.4 to 20.3% during 1.4 Vol% isoflurane; the percentage of values of 0 mmHg increased from 2.4 to 9.8% during 1.4 Vol.%. No changes of these parameters were detected in the control animals during the 3-h observation period.(ABSTRACT TRUNCATED AT 250 WORDS)

A retrospective analysis of 3 year's experience of an interdisciplinary approach to gallstone disease including shock-waves

1206 patients with gallstone disease were treated between January 1985 and December 1987, using an interdisciplinary concept that included surgery, endoscopic sphincterotomy, and extracorporeal shock-wave lithotripsy (ESWL). Twenty-five per cent of the patients who were admitted for gallbladder stones were treated by ESWL, whereas 75% underwent surgery. Mortality of elective treatment for gallbladder stones amounted to 0.25% (0.4% in surgery, 0% in ESWL). Postoperative complication rate was low (4.2% in surgery, 7.0% in ESWL). After ESWL treatment, 80% of the patients were free of stones after a follow-up period of 1 year. Recurrence rate in these patients amounted up to 10%; in seven of 70 patients, mean follow-up period was 6 months after complete disappearance of stones. Twenty-seven per cent of all patients who were admitted for bile duct stones underwent surgery, whereas in the other 73%, calculi were removed via endoscopy. ESWL treatment was used additionally, if necessary. Fragments were left behind in three of 75 patients (4.0%) after surgical treatment, and in 7 of 200 patients (3.5%) after endoscopic and ESWL treatment, respectively. In the latter group, three patients (1.5%) required an additional operation. There were no deaths in either of the groups. The use of ESWL for treatment of gallbladder stones needs to be evaluated in long-term follow-up studies. Thus far, surgery remains the dominating method. Endoscopic procedures, eventually combined with ESWL, represent the preferred treatment for patients with bile duct stones.

Changes of immunoglobulin concentrations in the bile after proximal gastric vagotomy in rats

We have investigated the influence of proximal gastric vagotomy in rats (PGV) on the immunoglobulin concentration in the serum, bile, and intestinal fluid. Clear differences for serum IgA were noted: after PGV, rats had 1.5 mg/ml IgA in contrast to only 0.25 mg/ml in sham-operated controls. The other serum immunoglobulins remained unchanged. Bile immunoglobulins were elevated in PGV rats with regard to IgA, IgG1, IgG2a, and IgG2b. In addition, PGV rats had higher IgA, IgG1, IgG2a, and IgG2b concentrations in the intestinal fluid than controls. An explanation for these high Ig concentrations in the secretions might be the challenge by intestinal (microbial) antigens and, perhaps, mucosal inflammation with changes in the permeability. Indications for the former were the increase in the number of bacteria after PGV.

Application of a fibrinogen-thrombin-collagen-based hemostyptic agent in experimental injuries of liver and spleen

FTCH is a recently developed material which consists of a collagen fleece containing fibrinogen, thrombin, and aprotinin integrated into its surface. FTCH is highly effective in sealing of tissues and in establishing hemostasis. We evaluated FTCH in experimentally produced liver (n = 6) and splenic (n = 12) injuries in 18 adult mongrel dogs. The stability of the parenchymal seal of the splenic injuries was tested by splenic tissue pressure elevation after temporary ligation of the splenic vein. No breakthrough bleeding occurred up to a parenchymal pressure of 16.3 +/- 5 mm Hg. Complete hemostasis was easily achieved in all animals before closure. When the dogs were re-explored postoperatively at intervals of either 14 or 30 days, there was no gross evidence of recurrent bleeding. Histologic examinations demonstrated a partially regenerated capsule covering an unspecific fibrovascular granulation tissue and progressive resorption of FTCH without significant inflammatory response. We conclude the following: FTCH provides adequate hemostatic control of experimental liver and splenic injuries. FTCH has excellent tissue compatibility and can be applied easily and safely to hemorrhaging parenchymal wounds. It will not replace adequate surgical techniques, but could be useful as a quickly available and easily applicable hemostatic means in diffuse or acute bleeding of liver and spleen.