[Artificial neural networks as a psychiatric instrument]. / Artificiële neurale netwerken als psychiatrisch instrument.

BACKGROUND: Artificial intelligence (AI) has evolved enormously over the past decade and is increasingly being applied to a range of domains, including psychiatry. AI encompasses several modalities, including artificial neural networks (ANNs), referring to computer models partly based on the workings of the brain. ANNs have existed since the ’50s, but only became ‘mainstream’ since the 2010s. The fact that they are inspired by the workings of the brain raises the question of whether they can also be used to model the (dys)functioning of the brain. This question led to the advent of the research field ‘computational psychiatry’. AIM: This article aims at providing an accessible introduction to artificial neural networks, and potential applications hereof in contemporary psychiatric practice. METHOD: Literature review with some examples. RESULTS: In this article we try to outline with some concrete examples what artificial neural networks are and how they can be used to model mechanisms in the brain. We successively discuss ANNs as a model of the human visual system, as a model of prosopagnosia and as a model of auditory hallucinations and finally as a model of autism spectrum disorder. We also describe a number of limitations of this approach. CONCLUSION: A computer model that models the entire brain is challenging at present, but current models can help in testing hypotheses concerning possible mechanisms that give rise to a wide range of neuropsychiatric conditions.

[Simulator model for ultrasound-guided pericardiocentesis : Construction manual and practical evaluation]. / Simulationsmodell für ultraschallgestützte Perikardiozentese : Bauanleitung und praktische Evaluation.

An important challenge in learning ultrasound-assisted interventions, such as pericardiocentesis, is the navigation of the needle in a three-dimensional space on the basis of a two-dimensional image. In order to learn this in vitro realistic simulators are required. We manufactured a model which allows simulation of pericardiocentesis on the basis of ballistic gelatin (12.6%, 250 Bloom). Furthermore, the pericardiocentesis model was subjectively evaluated by 37 anesthetists in a pre-post design. The models used proved to be technically simple to manufacture, hard wearing and realistic. They are therefore regularly used in our hospitals to learn ultrasound-assisted interventions.

Format-dependent representations of symbolic and non-symbolic numbers in the human cortex as revealed by multi-voxel pattern analyses.

Neuroimaging studies in the last 20 years have tried to unravel the neural correlates of number processing across formats in humans and non-human primates. Results point to the intraparietal sulcus as the core area for an abstract representation of numerical quantity. On the other hand, there exist a variety of behavioral and neuroimaging data that are difficult to reconcile with the existence of such an abstract representation. In this study, we addressed this issue by applying multi-voxel pattern analysis (MVPA) to functional Magnetic Resonance Imaging (fMRI) data to unravel the neural representations of symbolic (digits) and non-symbolic (dots) numbers and their possible overlap on three different spatial scales (entire lobules, smaller regions of interest and a searchlight analysis with 2-voxel radius). Results showed that numbers in both formats are decodable in occipital, frontal, temporal and parietal regions. However, there were no overlapping representations between dots and digits on any of the spatial scales. These data suggest that the human brain does not contain an abstract representation of numerical magnitude.

Decline of the McCollough effect by orientation-specific post-adaptation exposure to achromatic gratings.

The McCollough effect is a contingent color after effect induced by adapting to colored gratings for several minutes. It has been demonstrated that a long-lasting adaptation effect such as the McCollough effect can be diminished by exposure to achromatic versions of the induction stimuli. However, the orientation specificity of this effect of post-adaptation exposure is not known. Here we report the findings from two experiments conducted to determine the influence of achromatic gratings and their orientation on the strength of the McCollough effect. After adaptation to the McCollough stimuli, participants were exposed to achromatic gratings or to one of two control conditions (either waiting in the dark or different-orientation achromatic gratings). Results suggest a significant decline of the McCollough effect after perceiving achromatic gratings in comparison to waiting in the dark or being exposed to different-orientation achromatic gratings. Thus, the effect of post-adaptation exposure to achromatic gratings is orientation specific.

Effects of perceptual learning in visual backward masking on the responses of macaque inferior temporal neurons.

Learning is critical for fast and efficient object recognition in primates. To understand the neuronal correlates of behavioral improvements due to training, we recorded the responses of single neurons in the inferior temporal (IT) cortex of monkeys that were trained to recognize briefly presented, backward-masked objects. First we investigated training effects that are specific to the objects shown during training and that do not transfer to untrained objects. Only one of two monkeys tested showed object-specific training effects at the behavioral level, and only this monkey showed a transient object-specific increase in object selectivity for trained compared with untrained backward-masked objects. However, in each monkey a substantial part of the training effect transferred to untrained objects. To investigate the neural correlates of these object-independent training effects, we compared the neural responses to masked objects in trained monkeys to the responses in untrained monkeys. Training was associated with a reduction of the responses to the irrelevant masking patterns. These findings suggest that extensive training in recognizing backward-masked objects results in neural changes that reduce IT responses to the interfering irrelevant masking patterns and enhance the processing of the relevant objects.

Prospective cross-sectional study of haemostatic factors in patients with and without coronary artery disease.

The role of haemostatic factors for arterial thrombosis, especially the prevalence of activated protein C (APC) resistance in patients with coronary artery disease (CAD), is controversial. Between November 1996 and August 1997, 665 patients were analyzed. Diagnosis of CAD was confirmed by coronary angiography, exclusion of CAD was accepted in the presence of negative stress testing or a negative coronary angiography. CAD was present in 370 (56%) and excluded in 295 (44%) patients. Patients with CAD were older (64 +/- 9.2 versus 57.7 +/- 16 years; P

Inferotemporal neurons represent low-dimensional configurations of parameterized shapes.

Behavioral studies with parameterized shapes have shown that the similarities among these complex stimuli can be represented using a low number of dimensions. Using psychophysical measurements and single-cell recordings in macaque inferotemporal (IT) cortex, we found an agreement between low-dimensional parametric configurations of shapes and the representation of shape similarity at the behavioral and neuronal level. The shape configurations, computed from both the perceived and neuron-based similarities, revealed a low number of dimensions and contained the same stimulus order as the parametric configurations. However, at a metric level, the behavioral and neural representations deviated consistently from the parametric configurations. These findings suggest an ordinally faithful but metrically biased representation of shape similarity in IT.

Can neuroimaging really tell us what the human brain is doing? The relevance of indirect measures of population activity.

Neuroimaging studies using positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) give an indication towards the localization of mental representations and processes in the human brain. It is not clear to what extent such global measures of neuronal activity, pooling across large populations of neurons, can reveal how certain computations are implemented by the neurons in such population ('computational neuroimaging'). Population activity is related tightly to single-cell activity when all neurons in the population have similar response properties. We describe some evidence from single-cell recordings in monkeys that indicates that neurons with similar response properties are not scattered randomly throughout the visual cortex. Notwithstanding this clustering, populations of nearby neurons are still rather heterogeneous, requiring some prudence in deriving single-cell response properties from population activity. The following review of recent neuroimaging studies of the visual system describes to what degree inferences about computations and representations can be drawn from these studies.

Visual object categorisation at distinct levels of abstraction: a new stimulus set.

We designed a new stimulus set with 269 line drawings of everyday artifacts and animals. The stimulus set contains several typical exemplars from a sample of 25 basic-level categories. We determined to what extent these stimuli were named at the basic level and at a more subordinate level. An additional experiment showed the validity of this calibration: typicality ratings were correlated significantly with the level of naming. In a final experiment we found that this effect depends largely on the global configuration of a stimulus as it was still apparent with degraded images obtained by locally shifting small fragments of the drawings.

Spatial sensitivity of macaque inferior temporal neurons.

Recent findings in dorsal visual stream areas and computational work raise the question whether neurons at the end station of the ventral visual stream can code for stimulus position. The authors provide the first detailed, quantitative data on the spatial sensitivity of neurons in the anterior part of the inferior temporal cortex (area TE) in awake, fixating monkeys. They observed a large variation in receptive field (RF) size (ranging from 2.8 degrees to 26 degrees ). TE neurons differed in their optimal position, with a bias toward the foveal position. Moreover, the RF profiles of most TE neurons could be fitted well with a two-dimensional Gaussian function. Most neurons had only one region of high sensitivity and showed a smooth decline in sensitivity toward more distal positions. In addition, the authors investigated some of the possible determinants of such spatial sensitivity. First, testing with low-pass filtered versions of the stimuli revealed that the general preference for the foveal position and the size of the RFs was not due simply to TE neurons receiving input with a lower spatial resolution at more eccentric positions. The foveal position was still preferred after intense low-pass filtering. Second, although an increase in stimulus size consistently broadened spatial sensitivity profiles, it did not change the qualitative features of these profiles. Moreover, size selectivity of TE neurons was generally position invariant. Overall, the results suggest that TE neurons can code for the position of stimuli in the central region of the visual field.

The representation of shape in the context of visual object categorization tasks.

To investigate the role of human fusiform gyrus in shape processing, we determined the effect of shape degradation on BOLD contrast in this region with fMRI during three tasks requiring subjects to determine either whether two successively presented nonsense shapes had the same global orientation (OR task); whether two successively presented meaningful objects belonged to the same basic level category (CAT task); or whether two successively presented objects represented the same exemplar of a category (EX task). On the behavioral level, shape degradation by locally shifting the pixels constituting the lines of stimuli had no effect on performance in the OR task, while it was detrimental to performance in the CAT and EX tasks. In comparison to the OR task, both the CAT and EX tasks were associated with activations in the occipitotemporal and parietal cortex. When shape degradation was applied, activation in the middle fusiform gyrus was reduced in all tasks. The occurrence of this effect in the OR task indicates that it is independent of memory representations. The persistence of the effect in both tasks that showed a behavioral effect of degradation suggests that it does not reflect the amount of shape processing performed on the stimuli, but rather the specificity of the final perceptual representation that can be built from the shape information that is available. Other studies have shown effects of stimulus familiarity and task requirements in the fusiform gyrus, suggesting that there is no need to assume different modules for perceptual representation and representation in memory.

Measurement of antithrombin activity by thrombin-based and by factor Xa-based chromogenic substrate assays.

Functionally active antithrombin can be quantified by chromogenic substrate assays utilizing the heparin cofactor activity of antithrombin and the inhibition rates of thrombin or of activated factor X (FXa). Thrombin-based assays but not FXa-based assays may overestimate the antithrombin activity due to their sensitivity toward heparin cofactor II. We focused on the question whether an overestimation of antithrombin activity by thrombin-based assays involves the risk of misdiagnosing antithrombin-deficient individuals as being non-deficient. We determined antithrombin using two thrombin-based assays and one FXa-based assay in 27 plasma samples from patients with acquired antithrombin deficiency spiked with lepirudin, in antithrombin-deficient plasma and in mixtures of antithrombin-deficient plasma and normal plasma. We also measured antithrombin in healthy subjects, in patients with inherited and acquired antithrombin deficiency and in patients under high-dose heparin treatment. At therapeutic final concentrations of lepirudin, antithrombin activities were considerably overestimated by the thrombin-based assays but not by the FXa-based assay. The residual antithrombin activities in antithrombin-deficient plasma determined by the thrombin-based assays were markedly higher than the corresponding values obtained with the FXa-based assay. The thrombin-based assays also overestimated antithrombin activity in patients under high-dose heparin. However, the degree of overestimation in the range between 50 and 100 IU/dl was too low to misidentify individuals with inherited or acquired antithrombin deficiency as normal. We conclude that functionally active antithrombin can be reliably determined using FXa-based chromogenic substrate assays in all settings examined. Thrombin-based assays must not be used in patients under treatment with hirudin or other direct thrombin inhibitors.

The influence of citrate concentration on the quality of plasma obtained by automated plasmapheresis: a prospective study.

BACKGROUND: There is a need for more comprehensive work dealing with the quality of plasma collected by automated plasmapheresis using different final concentrations of citrate anticoagulant. A prospective study was performed to examine the influence of three concentrations of sodium citrate on the levels of clotting factors and markers of activated hemostasis and fibrinolysis. STUDY DESIGN AND METHODS: Fifty-one experienced plasma donors were recruited for subsequent 750-mL plasmapheresis procedures using 4-percent (wt/vol) sodium citrate. Anticoagulant-to-blood ratios of 1:16.6, 1:14.2, and 1:12.5 were used, corresponding to sodium citrate concentrations of 6 percent, 7 percent, and 8 percent (vol/vol), respectively. Between two plasmapheresis procedures, there was a washout period of 7 days. Determinations were made of the plasma levels of fibrinogen and factors V, VII, VIII, and IX, as well as antithrombin, tissue-type plasminogen activator, and several markers of activated hemostasis and fibrinolysis: activated factor VII, prothrombin splits products, D-dimers, and beta-thromboglobulin. RESULTS: The plasma samples anticoagulated with 6-percent citrate contained significantly higher levels of factors V, VIII, and IX than the samples anticoagulated with 8-percent citrate (p<0.0001, p< or =0.0001 and p = 0.009, respectively). The citrate concentration had no influence on the levels of fibrinogen, factor VII, antithrombin, or tissue-type plasminogen activator. There was no evidence that the plasma samples containing lower citrate concentrations were more prone to activation of hemostasis or fibrinolysis. CONCLUSION: A reduction in the final citrate concentration of plasma collected by automated plasmapheresis results in higher yields of factors V, VIII, and IX without activation of hemostasis. More comprehensive studies should confirm previous work dealing with the establishment of the lowest citrate concentration acceptable in plasma used as therapeutic fresh-frozen plasma or as starting material for the manufacture of plasma derivatives.

In vitro characterization of solvent/detergent-treated human plasma and of quarantine fresh frozen plasma.

BACKGROUND AND OBJECTIVES: Plasma pools, solvent/detergent(S/D)-treated plasma produced from plasma pools, and single donor fresh frozen plasma that had been quarantined for at least 6 months (QFFP) differ in their composition regarding clotting factors, inhibitors and other important plasma proteins. There are poor data concerning stability of important clotting factors after thawing of frozen plasma units. MATERIALS AND METHODS: 12 plasma pools, 12 batches of S/D plasma produced from these respective plasma pools, and 12 units of QFFP were extensively analysed. The stability of fibrinogen and factors V, VII, and VIII after thawing, storage at room temperature and at +4 degrees C was also examined. RESULTS: We extensively analysed plasma pools before and after solvent/detergent treatment as well as quarantined single donor plasma units for parameters of coagulation and fibrinolysis. After the S/D step, all clotting factor activities and the activities of most inhibitors and other plasma proteins were in the normal range in all batches. Protein S and plasmin inhibitor activities decreased by 35% and 76%, respectively. S/D treatment partly activated factor VII (FVII). However, there were no marked increases of other markers of activated hemostasis. The interindividual variations of all proteins analysed were significantly lower in the S/D plasmas than in the single donor plasma units. An 8-hour storage lead to a marked decrease of FVIII activity, whereas there was no significant influence on fibrinogen and factors V and VII. CONCLUSIONS: There are no critical reductions of the activities of clotting factors, inhibitors, and other important plasma proteins due to S/D treatment. Efficacy and safety of S/D plasma is not hampered by reduced activities of protein S and plasmin inhibitor. Dosage calculation and the evaluation of clinical response is simplified by usage of the more standardized S/D plasma compared to QFFP.

Measurement of factor VII and of activated factor VII in healthy individuals and in prothrombin complex concentrates.

We established reference ranges for factor VII clotting activity (FVII:C), factor VII amidolytic activity (FVII:AM), and activated factor VII (FVIIa) in 102 healthy individuals. The reference ranges were 65-160 U/100 ml, 70-165 U/100 ml, and 30-170 mU/ml, respectively (2.5 and 97.5 percentiles). Freezing and thawing of the plasma samples had no influence on the assay results. Due to the small sample size, the results were not influenced by gender, age, smoking habits, and oral contraceptive use. The plasma levels of FVII:C, FVII:AM, and FVIIa were significantly correlated with each other. The significant correlation between FVIIa and FVII:AM indicates that FVIIa is not completely independent of circulating FVII mass. There was also a significant, though weak, correlation between FVIIa and FVII:C/FVII:AM ratios. Sixteen batches of prothrombin complex concentrates (PCC) from 3 manufacturers were also analysed. FVIIa could be detected in all preparations, with considerable variations from batch to batch. In contrast to the results obtained in plasma from normal individuals, there was a close correlation between FVIIa and FVII:C/FVII:AM ratios. The preparations could be characterized by their FVII and FVIIa potencies and by their FVII:C/FVII:AM ratios. In PCC, FVII:C was very strongly correlated with FVIIa, whereas no significant correlation was observed between FVII:AM and FVII:C and between FVII:AM and FVIIa, respectively. These results demonstrate that the FVII:C assay used is sensitive for detecting FVIIa. Thus, we cannot confirm that FVIIa sensitivity of one-stage clotting assays for FVII:C is low when a rabbit thromboplastin and a non-adsorbed FVII-deficient plasma is used.

Factor VII and activated-factor-VII content of prothrombin complex concentrates. The PCC Study Group.

BACKGROUND AND OBJECTIVES: The aim of this study was to determine the potencies of factor VII (FVII) and of activated FVII (FVIIa) in prothrombin complex concentrates (PCC). MATERIALS AND METHODS: We examined 56 lots of PCC from 5 manufacturers. Three brands were licensed preparations, and 1 product series had been involved in thromboembolic complications. FVII and FVIIa were measured using a two-stage amidolytic assay and a specific clotting assay, respectively. We also quantified FVII clotting activity by a one-stage assay reflecting a mixture of FVII zymogen and FVIIa. RESULTS: All PCC contained substantial amounts of FVII, and FVIIa could be detected in all lots. There were marked differences between manufacturers and some significant variabilities between batches. The two lots involved in thromboembolic events contained considerably more FVIIa than the PCC still licensed. The lowest FVIIa potencies were observed in an experimental product series, indicating that PCC can be produced without activation of FVII during the manufacturing process. CONCLUSION: FVIIa is present in all PCC containing FVII. High FVIIa potencies may contribute to the thrombogenic potential of these preparations, and determination of FVIIa potencies should be included in the in vitro characterization of PCC.

Synthesis, intracellular processing and secretion of thrombospondin in human endothelial cells.

The biosynthesis of thrombospondin, a glycoprotein first described in platelets, has been studied in human endothelial cells. This glycoprotein has a molecular mass of 450 kDa. It is secreted and incorporated into the extracellular matrix of several cell types in culture. Pulse-chase experiments with [3H]leucine were performed and the synthesis and secretion of the glycoprotein was studied by immunoprecipitation and sodium dodecyl sulfate/polyacrylamide gel electrophoresis. The results of these experiments show that the three subunits of thrombospondin are identical in molecular mass. During synthesis there is a small but significant increase in molecular mass within 20 min after pulse labeling. The early form of thrombospondin is sensitive to endoglucosaminidase H treatment, indicating that a transformation of the oligosaccharide structures from 'high-mannose' to 'complex' structures takes place. Within 60 min after synthesis only the mature form of the glycoprotein is secreted into the medium. In the presence of tunicamycin, an inhibitor of N-glycosylation, there is a reduction in molecular mass of the subunit from 165 kDa to 155 kDa. Pulse-chase experiments in the presence of tunicamycin supported the conclusion that the carbohydrate part is processed during biosynthesis. Inhibition of glycosylation had a pronounced effect on the secretion of thrombospondin. The decreased occurrence of thrombospondin in the culture medium seemed to be due to a high intracellular degradation rate of unglycosylated thrombospondin. Characterization of the glycopeptide structures of thrombospondin metabolically labeled with [3H]mannose by Bio-Gel P-4 and concanavalin-A-Sepharose column chromatography revealed that the oligosaccharide structures of the cellular and secreted forms of thrombospondin differ in their composition.

A double antibody sandwich microELISA for measuring human platelet factor 4.

With the aim of investigating patients with a higher risk of thrombosis we developed a method for determining platelet factor 4 (PF4) in human plasma. Using the double antibody sandwich ELISA technique we set up a test system that allows the determination of PF4 concentrations in plasma samples from 1 to 100 ng/ml. This method is more sensitive and as specific as commercially available RIA kits, but has the advantage of being cheaper and less time consuming. Furthermore the ELISA does not require radioactive materials. The complete reaction is carried out in microtiter wells, and an ELISA-reader connected to an Apple computer does all the calculations needed for quantitative measurements.

[Carbon-replica-cytochemistry--a method for localization of cell surface components (author's transl)]. / Kohlefilmcytochemie--ein Verfahren zur Lokalisation von zellmenbran-assoziierten Stoffen.

A method was used which combines immunocytochemical techniques with surface replication by carbon films in order to localize fibronectin and N-acetyl-beta-hexosaminidase on cell surfaces of human fibroblasts in vitro. Similar results were received combining the replica techniques with conventional enzyme-cytochemistry.