Professor Juanita Todd

The number of women employed in STEM in Australia is increasing, however, they continue to remain underrepresented in most industries. A significant corpus of literature on female underrepresentation has emerged in the past 20 years, however, many of those studies focus on educational access and retention and not many look at the lived experiences of women after they have left higher education. In this article, we take a different stance and explore the heterogeneous experiences of female STEM professionals in regional Australia. Through the qualitative analysis of 25 interviews, we learn what women have endured, accepted, and valued on their individual STEM journeys. While these journeys are often quite different, our interviewees independently reported having experienced similar societal prejudices and possessing similar personality traits. Our data reveals that resilience and determination proved vital for these women, as did a strong early interest in STEM. Our interviews also unearth issues in which women¿s opinions are fiercely divided, such as whether positive discrimination has been a barrier or an enabler for their careers. Based on what we have learnt from their accounts, we argue that these women have ¿survived¿ their work environments despite structural barriers, only due to their determination, resilience and fervent interest.

The purpose of this review is to provide an accessible exploration of key considerations of lateralization in speech and non-speech perception using clear and defined language. From these considerations, the primary arguments for each side of the linguistics versus acoustics debate are outlined and explored in context of emerging integrative theories. This theoretical approach entails a perspective that linguistic and acoustic features differentially contribute to leftward bias, depending on the given context. Such contextual factors include stimulus parameters and variables of stimulus presentation (e.g., noise/silence and monaural/binaural) and variances in individuals (sex, handedness, age, and behavioural ability). Discussion of these factors and their interaction is also aimed towards providing an outline of variables that require consideration when developing and reviewing methodology of acoustic and linguistic processing laterality studies. Thus, there are three primary aims in the present paper: (1) to provide the reader with key theoretical perspectives from the acoustics/linguistics debate and a synthesis of the two viewpoints, (2) to highlight key caveats for generalizing findings regarding predominant models of speech laterality, and (3) to provide a practical guide for methodological control using predominant behavioural measures (i.e., gap detection and dichotic listening tasks) and/or neurophysiological measures (i.e., mismatch negativity) of speech laterality.

Mismatch Negativity (MMN) to pattern deviations reveals exquisite pattern detection ability in the brain. MMN amplitude is proposed to be precision-weighted, being inversely proportional to variability within a patterned sound sequence. Two experiments were conducted to determine whether pattern variability, shown to influence MMN to simple pattern deviance, also extends to MMN elicited to abstract pattern deviants. Participants were presented with 3-tone triplet sequences that were defined by regular frequency ascendance with adjacent (A<B<C) or non-adjacent (A<C) dependency. The triplets were defined by an abstract pattern in that the starting frequency of A roamed randomly between 500¿3700 Hz. Using variants of these sequences over two studies the results show that MMN was elicited to rare A > C deviants for adjacent and non-adjacent dependencies, was smaller for the latter, was impervious to variance in tone loudness, but showed prolonged sensitivity to the level of variability at sequence onset.

The concept of perceptual inferences taking place over multiple timescales simultaneously raises questions about how the brain can balance the demands of remaining sensitive to local rarity while utilising more global longer-term predictability to modulate cortical responses. In the present study auditory evoked potentials to four presentations of the same sound sequence containing predictable structure on a local (milliseconds to seconds) and more global (many minutes) timescales were recorded. The results from 33 participants are used to demonstrate that predictions about both local (internal predictive models) and global (meta-models that define expected precisions associated with familiar internal model states) regularities are formed. The study exposes more local context-based modulations of the P1 but more global order-based modulations of the auditory evoked N2 components. The results are discussed in terms of theoretical links advocating that uncertainty at multiple timescales could lead to differential component modulations, and the importance of considering the broader learning context in auditory evoked potential studies.

Rodent models play a significant role in understanding disease mechanisms and the screening of new treatments. With regard to psychiatric disorders such as schizophrenia, however, it is difficult to replicate the human symptoms in rodents because these symptoms are often either ¿uniquely human¿ or are only conveyed via self-report. There is a growing interest in rodent mismatch responses (MMRs) as a translatable ¿biomarker¿ for disorders such as schizophrenia. In this review, we will summarize the attributes of human MMN, and discuss the scope of exploring the attributes of human MMN in rodents. Here, we examine how reliably MMRs that are measured in rats mimic human attributes, and present original data examining whether manipulations of stimulus conditions known to modulate human MMN, do the same for rat MMRs. Using surgically-implanted epidural electroencephalographic electrodes and wireless telemetry in freely-moving rats, we observed human-like modulations of MMRs, namely that larger MMRs were elicited to unexpected (deviant) stimuli that a) had a larger change in pitch compared to the expected (standard) stimulus, b) were less frequently presented (lower probability), and c) had no jitter (stable stimulus onset asynchrony) compared to high jitter. Overall, these findings contribute to the mounting evidence for rat MMRs as a good analogue of human MMN, bolstering the development of a novel approach in future to validate the preclinical models based on a translatable biomarker, MMN.

Our understanding of the sensory environment is contextualized on the basis of prior experience. Measurement of auditory ERPs provides insight into automatic processes that contextualize the relevance of sound as a function of how sequences change over time. However, task-independent exposure to sound has revealed that strong first impressions exert a lasting impact on how the relevance of sound is contextualized. Dynamic causal modeling was applied to auditory ERPs collected during presentation of alternating pattern sequences. A local regularity (a rare p = .125 vs. common p = .875 sound) alternated to create a longer timescale regularity (sound probabilities alternated regularly creating a predictable block length), and the longer timescale regularity changed halfway through the sequence (the regular block length became shorter or lon-ger). Predictions should be revised for local patterns when blocks alternated and for longer patterning when the block length changed. Dynamic causal modeling revealed an overall higher precision for the error signal to the rare sound in the first block type, consistent with the first impression. The connectivity changes in response to errors within the underlying neural network were also different for the two blocks with significantly more revision of predictions in the arrangement that violated the first impression. Furthermore, the effects of block length change suggested errors within the first block type exerted more influence on the updating of longer timescale predictions. These observations support the hypothesis that automatic sequential learning creates a high-precision context (first impression) that impacts learning rates and updates to those learning rates when predictions arising from that context are violated. The results further evidence automatic pattern learning over multiple timescales simultaneously, even during task-independent passive exposure to sound.

One of the most robust electrophysiological features of schizophrenia is reduced mismatch negativity, a component of the event related potential (ERP) induced by rare and unexpected stimuli in an otherwise regular pattern. Emerging evidence suggests that mismatch negativity (MMN) is not the only ERP index of deviance detection in the mammalian brain and that sensitivity to deviant sounds in a regular background can be observed at earlier latencies in both the human and rodent brain. Pharmacological studies in humans and rodents have previously found that MMN reductions similar to those seen in schizophrenia can be elicited by N-methyl-D-aspartate (NMDA) receptor antagonism, an observation in agreement with the hypothesised role of NMDA receptor hypofunction in schizophrenia pathogenesis. However, it is not known how NMDA receptor antagonism affects early deviance detection responses. Here, we show that NMDA antagonism impacts both early and late deviance detection responses. By recording EEG in awake, freely-moving rats in a drug-free condition and after varying doses of NMDA receptor antagonist MK-801, we found the hypothesised reduction of deviance detection for a late, negative potential (N55). However, the amplitude of an early component, P13, as well as deviance detection evident in the same component, were increased by NMDA receptor antagonism. These findings indicate that late deviance detection in rats is similar to human MMN, but the surprising effect of MK-801 in increasing ERP amplitudes as well as deviance detection at earlier latencies suggests that future studies in humans should examine ERPs over early latencies in schizophrenia and after NMDA antagonism.

Time, or more specifically temporal structure, is a critical variable in understanding how the auditory system uses acoustic patterns to predict input, and to filter events based on their relevance. A key index of this filtering process is the auditory evoked potential component known as mismatch negativity or MMN. In this paper we review findings of smaller MMN in schizophrenia through the lens of time as an influential contextual variable. More specifically, we review studies that show how MMN to a locally rare pattern-deviation is modulated by the longer-term context in which it occurs. Empirical data is presented from a non-clinical sample confirming that the absence of a stable higher-order structure to sound sequences alters the way MMN amplitude changes over time. This result is discussed in relation to how hierarchical pattern learning might enrich our understanding of how and why MMN amplitude modulation is disrupted in schizophrenia.

The existence and updating of ¿sensory beliefs¿ or internal models can be studied using auditory evoked potentials (AEPs) when there is some form of predictable pattern in sound. Internal models are proposed to enable predictions about the most likely next sound-activation-state leading to small AEPs to standard sounds matching model predictions, and larger AEPs to sound that deviate. Internal models are precision-weighted with the standard-deviant difference being largest when precision is high (variability is low). Here we expose how order-effects determine whether a change in variability impacts model-precision estimates. Thirty participants heard 3000 t (30 ms standard p = 0.90 and 60 ms deviant p = 0.10) that either moved from a more precise stimulus onset asymmetry (n = 15, first 1000 tones 500 ms ± 10 ms) to a more variable one (n = 15, subsequent 2000 tones 500 ms ± 200 ms) or from variable (first 1000 t) to more precise (subsequent 2000 t). AEPs were equivalent between groups for the first 1000 tones but differed dramatically in the face of timing changes. Where timing precision decreased, the standard-deviant difference was impervious to the change but where precision increased, the standard-deviant difference increased dramatically after the timing change signalling a transient increase in model precision that subsided over the final 1000 tones. The results support contemporary models proposing that updates to an active internal model will be a function of the quality of the evidence upon which it has been built and the information value of subsequent errors in improving the predictive success of the active model.

Maternal exposure to infectious agents during gestation has been identified as a significant risk factor for schizophrenia. Using a mouse model, past work has demonstrated that the gestational timing of the immune-activating event can impact the behavioural phenotype and expression of dopaminergic and glutamatergic neurotransmission markers in the offspring. In order to determine the inter-species generality of this effect to rats, another commonly used model species, the current study investigated the impact of a viral mimetic Poly (I:C) at either an early (gestational day 10) or late (gestational day 19) time-point on schizophrenia-related behaviour and neurotransmitter receptor expression in rat offspring. Exposure to Poly (I:C) in late, but not early, gestation resulted in transient impairments in working memory. In addition, male rats exposed to maternal immune activation (MIA) in either early or late gestation exhibited sensorimotor gating deficits. Conversely, neither early nor late MIA exposure altered locomotor responses to MK-801 or amphetamine. In addition, increased dopamine 1 receptor mRNA levels were found in the nucleus accumbens of male rats exposed to early gestational MIA. The findings from this study diverge somewhat from previous findings in mice with MIA exposure, which were often found to exhibit a more comprehensive spectrum of schizophrenia-like phenotypes in both males and females, indicating potential differences in the neurodevelopmental vulnerability to MIA exposure in the rat with regards to schizophrenia related changes.

Objectives: People living with a psychotic illness have higher rates of cigarette smoking and face unique barriers to quitting compared to the general population. We examined whether self-reported reasons for smoking are useful predictors of successful quit attempts among people with psychosis. Methods: As part of a randomized controlled trial addressing smoking and cardiovascular disease risk behaviors among people with psychosis, self-reported reasons for smoking were assessed at baseline (n = 235), 15 weeks (n = 151), and 12 months (n = 139). Three factors from the Reasons for Smoking Questionnaire (Coping, Physiological, and Stimulation/Activation) were entered into a model to predict short- and long-term abstinence. The relationship between these factors and mental health symptoms were also assessed. Results: Participants scoring higher on the Stimulation/Activation factor (control of weight, enjoyment, concentration, and ¿peps me up¿) at baseline were just less than half as likely to be abstinent at 15 weeks. Female participants were five times more likely to abstinent at 15 weeks, and those with a higher global functioning at baseline were 5% more likely to be abstinent. There was a positive correlation between changes over time in the Stimulation/Activation factor from baseline to 12-month follow-up and the Brief Psychiatric Rating Scale total score at 12-month follow-up. This indicates that increasingly higher endorsement of the factor was associated with more psychological symptoms. There was also a negative correlation between the change over time in the Stimulation/Activation factor and global functioning at 12 months, indicating that increasingly higher endorsement of the factor led to lower global assessment of functioning. Conclusions: The Stimulation/Activation factor may be particularly important to assess and address among smokers with psychosis. It is recommended that further research use the Reasons for Smoking Questionnaire among smokers with psychosis as a clinical tool to identify specific quit barriers. Further research into why females have higher smoking cessation rates in the short term and relapse prevention interventions seem worthy of further investigation.

Schizophrenia may be conceptualised using a dimensional approach to examine trait-like expression such as schizotypy within non-clinical populations to better understand pathophysiology. A candidate psychosis-risk marker, the auditory mismatch negativity (MMN) is thought to index the functionality of glutamatergic NMDA receptor mediated neurotransmission. Although the MMN is robustly reduced in patients with schizophrenia, the association between MMN and schizotypy in the general population is under-investigated. Thirty-five healthy participants completed the Schizotypal Personality Questionnaire (SPQ) and a multi-feature MMN paradigm (standards 82%, 50¿ms, 1000¿Hz, 80¿dB) with duration (100¿ms), frequency (1200¿Hz) and intensity (90¿dB) deviants (6% each). Spearman's correlations were used to explore the association between schizotypal personality traits and MMN amplitude. Few associations were identified between schizotypal traits and MMN. Higher Suspiciousness subscale scores tended to be correlated with larger frequency MMN amplitude. A median-split comparison of the sample on Suspiciousness scores showed larger MMN (irrespective of deviant condition) in the High compared to the Low Suspiciousness group. The trend-level association between MMN and Suspiciousness is in contrast to the robustly attenuated MMN amplitude observed in schizophrenia. Reductions in MMN may reflect a schizophrenia-disease state, whereas non-clinical schizotypy may not be subserved by similar neuropathology.

Although first-impressions are known to impact decision-making and to have prolonged effects on reasoning, it is less well known that the same type of rapidly formed assumptions can explain biases in automatic relevance filtering outside of deliberate behavior. This paper features two studies in which participants have been asked to ignore sequences of sound while focusing attention on a silent movie. The sequences consisted of blocks, each with a high-probability repetition interrupted by rare acoustic deviations (i.e., a sound of different pitch or duration). The probabilities of the two different sounds alternated across the concatenated blocks within the sequence (i.e., short-to-long and long-to-short). The sound probabilities are rapidly and automatically learned for each block and a perceptual inference is formed predicting the most likely characteristics of the upcoming sound. Deviations elicit a prediction-error signal known as mismatch negativity (MMN). Computational models of MMN generally assume that its elicitation is governed by transition statistics that define what sound attributes are most likely to follow the current sound. MMN amplitude reflects prediction confidence, which is derived from the stability of the current transition statistics. However, our prior research showed that MMN amplitude is modulated by a strong first-impression bias that outweighs transition statistics. Here we test the hypothesis that this bias can be attributed to assumptions about predictable vs. unpredictable nature of each tone within the first encountered context, which is weighted by the stability of that context. The results of Study 1 show that this bias is initially prevented if there is no 1:1 mapping between sound attributes and probability, but it returns once the auditory system determines which properties provide the highest predictive value. The results of Study 2 show that confidence in the first-impression bias drops if assumptions about the temporal stability of the transition-statistics are violated. Both studies provide compelling evidence that the auditory system extrapolates patterns on multiple timescales to adjust its response to prediction-errors, while profoundly distorting the effects of transition-statistics by the assumptions formed on the basis of first-impressions.

Prolonged heavy exposure to cannabis is associated with impaired cognition and brain functional and structural alterations. We recently reported attenuated mismatch negativity (MMN) and altered P50 sensory gating in chronic cannabis users. This study investigated the extent of brain functional recovery (indexed by MMN and P50) in chronic users after cessation of use. Eighteen ex-users (median 13.5 years prior regular use; median 3.5 years abstinence) and 18 nonusers completed (1) a multifeature oddball task with duration, frequency, and intensity deviants and (2) a P50 paired-click paradigm. Trend level smaller duration MMN amplitude and larger P50 ratios (indicative of poorer sensory gating) were observed in ex-users compared to controls. Poorer P50 gating correlated with prior duration of cannabis use. Duration of abstinence was positively correlated with duration MMN amplitude, even after controlling for age and duration of cannabis use. Impaired sensory gating and attenuated MMN amplitude tended to persist in ex-users after prolonged cessation of use, suggesting a lack of full recovery. An association with prolonged duration of prior cannabis use may indicate persistent cannabis-related alterations to P50 sensory gating. Greater reductions in MMN amplitude with increasing abstinence (positive correlation) may be related to either self-medication or an accelerated aging process.

Although the scientific community appears to know a lot about MMN, about its neural generators and the computational processes that underlie its generation, do we have sufficient knowledge to understand what causes the reduction of MMN amplitude in schizophrenia? Here we attempt to integrate the evidence presented in this series of papers for the special issue on MMN in schizophrenia together with evidence from other new relevant research and ask-what have we learnt? While MMN research was the purview for decades of psychophysiologists interested in event-related potentials derived from scalp recorded EEG, it is now part of mainstream neuroscience research attracting the interest of basic auditory neuroscientists, neurobiologists and computational modellers. The confluence of these developments together with increasing clinical research has certainly advanced our understanding of the causes of reduced MMN in schizophrenia as this integrative review attempts to demonstrate-but much remains to be learnt. Future advances will rely on the application of multiple methodologies and approaches in order to arrive at better understanding of the neurobiology of MMN and implications for schizophrenia.

The early detection of young people at-risk of developing a severe mental illness like schizophrenia offers the opportunity of introducing treatment earlier than currently possible. There is some evidence that early intervention improves prognosis and functional outcome, or even prevents the full clinical manifestation of the condition in some individuals. A key prerequisite to facilitate early intervention would be a biomarker that can reliably predict a transition to schizophrenia. A smaller event-related mismatch negativity (MMN) potential has emerged as one of the most robust psychophysiological finding in schizophrenia akin of a biomarker of the condition. More recent research further demonstrates that MMN, but also P3a amplitudes, are already reduced in the prodromal phase of illness. Several lines of pre-clinical and clinical research support this notion and are reviewed in this article together with current obstacles, which are still limiting the translation of MMN as a biomarker into clinical practice.

The mismatch negativity (MMN) is conceptualized as a confidence-weighted error signal elicited when a deviation violates the predicted next-state based on regularity. The mechanisms underpinning its generation remain contentious. Smaller MMN response is a robust finding in schizophrenia and reduced amplitude may implicate impairment in prediction-error signalling. An enriched understanding of factors that influence MMN size in healthy people is a prerequisite for translating the relevance of reduced MMN in schizophrenia. This paper features two studies designed to explore factors that impact MMN in healthy individuals. Study 1 confirms that MMN amplitude does not faithfully reflect transition statistics and is susceptible to order-driven bias. In study 2, we demonstrate that an order-driven bias remains despite repeated encounters with sound sequences. These data demonstrate that factors that impact on MMN size in non-clinical groups are not fully understood and that some mechanisms driving relevance filtering are likely influenced by 'top-down' expectations.

Smoking prevalence in schizophrenia is significantly elevated relative to other clinical and to non-clinical groups. The cognitive self-medication hypothesis attributes this to the beneficial effects of nicotine on illness-related cognitive deficits. Significant effects of nicotine have been observed on visual spatial working memory (VSWM), sustained attention (Continuous Performance Test - Identical Pairs; CPT-IP) and prepulse inhibition (PPI). It remains unclear whether these neurophysiological and neurocognitive effects of nicotine influence self-reported smoking motivation. To explore the relationship between nicotine effects on cognition and self-reported smoking motivation in schizophrenia and non-psychiatric control smokers, the impact of smoking abstinence and reinstatement was examined across three cognitive indices (VSWM, CPT-IP, PPI) and compared to self-reported smoking motivation (Modified Reasons for Smoking Scale revised to include 'cognitive motivators'). Cognitive function was assessed after 'typical' smoking and overnight abstinence. Schizophrenia smokers (but not controls) demonstrated significantly less error on the VSWM task in the smoking relative to abstinent condition. Control (but not schizophrenia) smokers, showed evidence of CPT-IP improvement in the smoking relative to abstinent condition. The overall profile of smoking motivation was comparable between groups. However, significant relationships between subjective and objective indices of smoking related cognitive change were observed for controls. Differential effects of nicotine on cognition have been hypothesised to influence the pattern and persistence of smoking in schizophrenia. These preliminary findings indicate that evidence for such effects is apparent even in small samples - particularly for VSWM. This is the first study to show that neurocognitive effects of smoking may influence self-reported smoking motivation.

The aim of this review is to explore the phenomenon of reduced mismatch negativity (MMN) in persons with schizophrenia and the possible relationship it has with attention impairments. In doing so we discuss (i) the prediction error account of MMN, (ii) reduced MMN as a faulty predictive processing system in persons with schizophrenia, (iii) the role of these systems in relevance filtering and attentional resource protection, (iv) attentional impairments in persons with schizophrenia, and (v) research that has explored MMN and attention in schizophrenia groups. Our review of the literature suggests that no study has appropriately examined the functional impact of smaller MMN in schizophrenia on the performance of a concurrent attention task. We conclude that future research should explore this notion further in the hope that it might embed MMN findings within outcomes of functional significance to individuals with the illness and those providing treatment. This article is part of a Special Issue entitled SI: Prediction and Attention.

Mismatch negativity (MMN) is a scalp-recorded electrical potential that occurs in humans in response to an auditory stimulus that defies previously established patterns of regularity. MMN amplitude is reduced in people with schizophrenia. In this study, we aimed to develop a robust and replicable rat model of MMN, as a platform for a more thorough understanding of the neurobiology underlying MMN. One of the major concerns for animal models of MMN is whether the rodent brain is capable of producing a human-like MMN, which is not a consequence of neural adaptation to repetitive stimuli. We therefore tested several methods that have been used to control for adaptation and differential exogenous responses to stimuli within the oddball paradigm. Epidural electroencephalographic electrodes were surgically implanted over different cortical locations in adult rats. Encephalographic data were recorded using wireless telemetry while the freely-moving rats were presented with auditory oddball stimuli to assess mismatch responses. Three control sequences were utilized: the flip-flop control was used to control for differential responses to the physical characteristics of standards and deviants; the many standards control was used to control for differential adaptation, as was the cascade control. Both adaptation and adaptation-independent deviance detection were observed for high frequency (pitch), but not low frequency deviants. In addition, the many standards control method was found to be the optimal method for observing both adaptation effects and adaptation-independent mismatch responses in rats. Inconclusive results arose from the cascade control design as it is not yet clear whether rats can encode the complex pattern present in the control sequence. These data contribute to a growing body of evidence supporting the hypothesis that rat brain is indeed capable of exhibiting human-like MMN, and that the rat model is a viable platform for the further investigation of the MMN and its associated neurobiology.

Repetitive patterning facilitates inferences about likely properties of sound to follow. Mismatch negativity (MMN) occurs when sound fails to match an inference. Smaller MMN in schizophrenia indexes deficient gain control (difference in utilizing a limited dynamic range). Although it is clear that this group has a lower limit to MMN size, this study addressed whether smaller MMN indicates impaired perceptual inference. MMN was elicited to four deviants in two sequences: one in which occurrence was random and one in which it was paired. Despite smaller MMN, persons with schizophrenia are equally able to reduce MMN size evoked by a deviant when its occurrence is cued. Results also expose alterations in the evoked response to repeated sounds that appear to be exacerbations of age-related amplitude decline. Since these anomalies impact the computed MMN, they highlight the need to identify all contributions to limits in gain control in schizophrenia. © 2014 Society for Psychophysiological Research.

Repetitious patterns enable the auditory system to form prediction models specifying the most likely characteristics of subsequent sounds. Pattern deviations elicit mismatch negativity (MMN), the amplitude of which is modulated by the size of the deviation and confidence in the model. Todd et al. (Neuropsychologia 49:3399-3405, 2011; J Neurophysiol 109:99-105, 2013) demonstrated that a multi-timescale sequence reveals a bias that profoundly distorts the impact of local sound statistics on the MMN amplitude. Two sounds alternate roles as repetitious "standard" and rare "deviant" rapidly (every 0.8 min) or slowly (every 2.4 min). The bias manifests as larger MMN to the sound first encountered as deviant in slow compared to fast changing sequences, but no difference for the sound first encountered as a standard. We propose that the bias is due to how Bayesian priors shape filters of sound relevance. By examining the time-course of change in MMN amplitude we show that the bias manifests immediately after roles change but rapidly disappears thereafter. The bias was reflected in the response to deviant sounds only (not in response to standards), consistent with precision estimates extracted from second order patterns modulating gain differentially for the two sounds. Evoked responses to deviants suggest that pattern extraction and reactivation of priors can operate over tens of minutes or longer. Both MMN and deviant responses establish that: (1) priors are defined by the most proximally encountered probability distribution when one exists but; (2) when no prior exists, one is instantiated by sequence onset characteristics; and (3) priors require context interruption to be updated. © 2013 Springer Science+Business Media.