Learning Representations from Heart Sound: A Comparative Study on Shallow and Deep Models.

Leveraging the power of artificial intelligence to facilitate an automatic analysis and monitoring of heart sounds has increasingly attracted tremendous efforts in the past decade. Nevertheless, lacking on standard open-access database made it difficult to maintain a sustainable and comparable research before the first release of the PhysioNet CinC Challenge Dataset. However, inconsistent standards on data collection, annotation, and partition are still restraining a fair and efficient comparison between different works. To this line, we introduced and benchmarked a first version of the Heart Sounds Shenzhen (HSS) corpus. Motivated and inspired by the previous works based on HSS, we redefined the tasks and make a comprehensive investigation on shallow and deep models in this study. First, we segmented the heart sound recording into shorter recordings (10 s), which makes it more similar to the human auscultation case. Second, we redefined the classification tasks. Besides using the 3 class categories (normal, moderate, and mild/severe) adopted in HSS, we added a binary classification task in this study, i.e., normal and abnormal. In this work, we provided detailed benchmarks based on both the classic machine learning and the state-of-the-art deep learning technologies, which are reproducible by using open-source toolkits. Last but not least, we analyzed the feature contributions of best performance achieved by the benchmark to make the results more convincing and interpretable.

Dawn of the Transformer Era in Speech Emotion Recognition: Closing the Valence Gap.

Recent advances in transformer-based architectures have shown promise in several machine learning tasks. In the audio domain, such architectures have been successfully utilised in the field of speech emotion recognition (SER). However, existing works have not evaluated the influence of model size and pre-training data on downstream performance, and have shown limited attention to generalisation, robustness, fairness, and efficiency. The present contribution conducts a thorough analysis of these aspects on several pre-trained variants of wav2vec 2.0 and HuBERT that we fine-tuned on the dimensions arousal, dominance, and valence of MSP-Podcast, while additionally using IEMOCAP and MOSI to test cross-corpus generalisation. To the best of our knowledge, we obtain the top performance for valence prediction without use of explicit linguistic information, with a concordance correlation coefficient (CCC) of. 638 on MSP-Podcast. Our investigations reveal that transformer-based architectures are more robust compared to a CNN-based baseline and fair with respect to gender groups, but not towards individual speakers. Finally, we show that their success on valence is based on implicit linguistic information, which explains why they perform on-par with recent multimodal approaches that explicitly utilise textual information. To make our findings reproducible, we release the best performing model to the community.

Perception and classification of emotions in nonsense speech: Humans versus machines.

This article contributes to a more adequate modelling of emotions encoded in speech, by addressing four fallacies prevalent in traditional affective computing: First, studies concentrate on few emotions and disregard all other ones ('closed world'). Second, studies use clean (lab) data or real-life ones but do not compare clean and noisy data in a comparable setting ('clean world'). Third, machine learning approaches need large amounts of data; however, their performance has not yet been assessed by systematically comparing different approaches and different sizes of databases ('small world'). Fourth, although human annotations of emotion constitute the basis for automatic classification, human perception and machine classification have not yet been compared on a strict basis ('one world'). Finally, we deal with the intrinsic ambiguities of emotions by interpreting the confusions between categories ('fuzzy world'). We use acted nonsense speech from the GEMEP corpus, emotional 'distractors' as categories not entailed in the test set, real-life noises that mask the clear recordings, and different sizes of the training set for machine learning. We show that machine learning based on state-of-the-art feature representations (wav2vec2) is able to mirror the main emotional categories ('pillars') present in perceptual emotional constellations even in degradated acoustic conditions.

Automatic vocalisation-based detection of fragile X syndrome and Rett syndrome.

Fragile X syndrome (FXS) and Rett syndrome (RTT) are developmental disorders currently not diagnosed before toddlerhood. Even though speech-language deficits are among the key symptoms of both conditions, little is known about infant vocalisation acoustics for an automatic earlier identification of affected individuals. To bridge this gap, we applied intelligent audio analysis methodology to a compact dataset of 4454 home-recorded vocalisations of 3 individuals with FXS and 3 individuals with RTT aged 6 to 11 months, as well as 6 age- and gender-matched typically developing controls (TD). On the basis of a standardised set of 88 acoustic features, we trained linear kernel support vector machines to evaluate the feasibility of automatic classification of (a) FXS vs TD, (b) RTT vs TD, (c) atypical development (FXS+RTT) vs TD, and (d) FXS vs RTT vs TD. In paradigms (a)-(c), all infants were correctly classified; in paradigm (d), 9 of 12 were so. Spectral/cepstral and energy-related features were most relevant for classification across all paradigms. Despite the small sample size, this study reveals new insights into early vocalisation characteristics in FXS and RTT, and provides technical underpinnings for a future earlier identification of affected individuals, enabling earlier intervention and family counselling.

Face mask recognition from audio: The MASC database and an overview on the mask challenge.

The sudden outbreak of COVID-19 has resulted in tough challenges for the field of biometrics due to its spread via physical contact, and the regulations of wearing face masks. Given these constraints, voice biometrics can offer a suitable contact-less biometric solution; they can benefit from models that classify whether a speaker is wearing a mask or not. This article reviews the Mask Sub-Challenge (MSC) of the INTERSPEECH 2020 COMputational PARalinguistics challengE (ComParE), which focused on the following classification task: Given an audio chunk of a speaker, classify whether the speaker is wearing a mask or not. First, we report the collection of the Mask Augsburg Speech Corpus (MASC) and the baseline approaches used to solve the problem, achieving a performance of 71.8 % Unweighted Average Recall (UAR). We then summarise the methodologies explored in the submitted and accepted papers that mainly used two common patterns: (i) phonetic-based audio features, or (ii) spectrogram representations of audio combined with Convolutional Neural Networks (CNNs) typically used in image processing. Most approaches enhance their models by adapting ensembles of different models and attempting to increase the size of the training data using various techniques. We review and discuss the results of the participants of this sub-challenge, where the winner scored a UAR of 80.1 % . Moreover, we present the results of fusing the approaches, leading to a UAR of 82.6 % . Finally, we present a smartphone app that can be used as a proof of concept demonstration to detect in real-time whether users are wearing a face mask; we also benchmark the run-time of the best models.

New Avenues in Audio Intelligence: Towards Holistic Real-life Audio Understanding.

Computer audition (i.e., intelligent audio) has made great strides in recent years; however, it is still far from achieving holistic hearing abilities, which more appropriately mimic human-like understanding. Within an audio scene, a human listener is quickly able to interpret layers of sound at a single time-point, with each layer varying in characteristics such as location, state, and trait. Currently, integrated machine listening approaches, on the other hand, will mainly recognise only single events. In this context, this contribution aims to provide key insights and approaches, which can be applied in computer audition to achieve the goal of a more holistic intelligent understanding system, as well as identifying challenges in reaching this goal. We firstly summarise the state-of-the-art in traditional signal-processing-based audio pre-processing and feature representation, as well as automated learning such as by deep neural networks. This concerns, in particular, audio interpretation, decomposition, understanding, as well as ontologisation. We then present an agent-based approach for integrating these concepts as a holistic audio understanding system. Based on this, concluding, avenues are given towards reaching the ambitious goal of 'holistic human-parity' machine listening abilities.

Computer Audition for Fighting the SARS-CoV-2 Corona Crisis-Introducing the Multitask Speech Corpus for COVID-19.

Computer audition (CA) has experienced a fast development in the past decades by leveraging advanced signal processing and machine learning techniques. In particular, for its noninvasive and ubiquitous character by nature, CA-based applications in healthcare have increasingly attracted attention in recent years. During the tough time of the global crisis caused by the coronavirus disease 2019 (COVID-19), scientists and engineers in data science have collaborated to think of novel ways in prevention, diagnosis, treatment, tracking, and management of this global pandemic. On the one hand, we have witnessed the power of 5G, Internet of Things, big data, computer vision, and artificial intelligence in applications of epidemiology modeling, drug and/or vaccine finding and designing, fast CT screening, and quarantine management. On the other hand, relevant studies in exploring the capacity of CA are extremely lacking and underestimated. To this end, we propose a novel multitask speech corpus for COVID-19 research usage. We collected 51 confirmed COVID-19 patients' in-the-wild speech data in Wuhan city, China. We define three main tasks in this corpus, i.e., three-category classification tasks for evaluating the physical and/or mental status of patients, i.e., sleep quality, fatigue, and anxiety. The benchmarks are given by using both classic machine learning methods and state-of-the-art deep learning techniques. We believe this study and corpus cannot only facilitate the ongoing research on using data science to fight against COVID-19, but also the monitoring of contagious diseases for general purpose.

SEWA DB: A Rich Database for Audio-Visual Emotion and Sentiment Research in the Wild.

Natural human-computer interaction and audio-visual human behaviour sensing systems, which would achieve robust performance in-the-wild are more needed than ever as digital devices are increasingly becoming an indispensable part of our life. Accurately annotated real-world data are the crux in devising such systems. However, existing databases usually consider controlled settings, low demographic variability, and a single task. In this paper, we introduce the SEWA database of more than 2,000 minutes of audio-visual data of 398 people coming from six cultures, 50 percent female, and uniformly spanning the age range of 18 to 65 years old. Subjects were recorded in two different contexts: while watching adverts and while discussing adverts in a video chat. The database includes rich annotations of the recordings in terms of facial landmarks, facial action units (FAU), various vocalisations, mirroring, and continuously valued valence, arousal, liking, agreement, and prototypic examples of (dis)liking. This database aims to be an extremely valuable resource for researchers in affective computing and automatic human sensing and is expected to push forward the research in human behaviour analysis, including cultural studies. Along with the database, we provide extensive baseline experiments for automatic FAU detection and automatic valence, arousal, and (dis)liking intensity estimation.

Can Machine Learning Assist Locating the Excitation of Snore Sound? A Review.

In the past three decades, snoring (affecting more than 30 % adults of the UK population) has been increasingly studied in the transdisciplinary research community involving medicine and engineering. Early work demonstrated that, the snore sound can carry important information about the status of the upper airway, which facilitates the development of non-invasive acoustic based approaches for diagnosing and screening of obstructive sleep apnoea and other sleep disorders. Nonetheless, there are more demands from clinical practice on finding methods to localise the snore sound's excitation rather than only detecting sleep disorders. In order to further the relevant studies and attract more attention, we provide a comprehensive review on the state-of-the-art techniques from machine learning to automatically classify snore sounds. First, we introduce the background and definition of the problem. Second, we illustrate the current work in detail and explain potential applications. Finally, we discuss the limitations and challenges in the snore sound classification task. Overall, our review provides a comprehensive guidance for researchers to contribute to this area.

Extension of the Mechanistic Tissue Distribution Model of Rodgers and Rowland by Systematic Incorporation of Lysosomal Trapping: Impact on Unbound Partition Coefficient and Volume of Distribution Predictions in the Rat.

Physiologically based pharmacokinetic modeling has become a standard tool to predict drug distribution in early stages of drug discovery; however, this does not currently encompass lysosomal trapping. For basic lipophilic compounds, lysosomal sequestration is known to potentially influence intracellular as well as tissue distribution. The aim of our research was to reliably predict the lysosomal drug content and ultimately integrate this mechanism into pharmacokinetic prediction models. First, we further validated our previously presented method to predict the lysosomal drug content (Schmitt et al., 2019) for a larger set of compounds (n = 41) showing a very good predictivity. Using the lysosomal marker lipid bis(monoacylglycero)phosphate, we estimated the lysosomal volume fraction for all major tissues in the rat, ranging from 0.03% for adipose up to 5.3% for spleen. The pH-driven lysosomal trapping was then estimated and fully integrated into the mechanistic distribution model published by Rodgers et al. (2005) Predictions of Kpu improved for all lysosome-rich tissues. For instance, Kpu increased for nicotine 4-fold (spleen) and 2-fold (lung and kidney) and for quinidine 1.8-fold (brain), although for most other drugs the effects were much less (≤7%). Overall, the effect was strongest for basic compounds with a lower lipophilicity, such as nicotine, for which the unbound volume of distribution at steady-state prediction changed from 1.34 to 1.58 l/kg. For more lipophilic (basic) compounds or those that already show strong interactions with acidic phospholipids, the additional contribution of lysosomal trapping was less pronounced. Nevertheless, lysosomal trapping will also affect intracellular distribution of such compounds. SIGNIFICANCE STATEMENT: The estimation of the lysosomal content in all body tissues facilitated the incorporation of lysosomal sequestration into a general physiologically based pharmacokinetic model, leading to improved predictions as well as elucidating its influence on tissue and subcellular distribution in the rat.

Quantitative molecular tissue atlas of Bis(monoacylglycero)phosphate and phosphatidylglycerol membrane lipids in rodent organs generated by methylation assisted high resolution mass spectrometry.

Bis(monoacylglycero)phosphate (BMP) and phosphatidylglycerol (PG) are structural isomeric phospholipids with very different properties and biological functions. Due to their isomeric nature, it has thus far been challenging to simultaneously quantify BMP and PG lipids in tissue samples by mass spectrometry. Therefore, we have developed a sensitive LC-MS/MS based approach with prior methylation derivatization that is able to handle large batches of samples. Using this high throughput platform, a simulated MS/MS database was established for confident lipid assignment. In this work, we have simultaneously identified and quantified BMP and PG lipid molecules in different body tissues of rats and mice. We report for the first time a quantitative molecular atlas of BMP and PG lipids for 14 different tissues and organs in Wistar rats, NMRI and CD1 mice. Organ- and species-specificity was analyzed and compared for both lipid molecule classes. A total of 34 BMP and 10 PG molecules were quantified, with PG concentrations being generally much higher across tissues than BMP, but BMP lipids showing a much higher molecular diversity between animal organs. The large diversity of the BMP lipids with regard to their abundance and molecular composition suggests distinct biological function(s) of the individual BMP molecules in different tissues and organs of body. Particularly high tissue levels of BMP were seen in spleen, lung, liver, kidney and small intestines, i.e. tissues that are known for their high abundance and/or activity level of lysosomes late and endosomes. Elevated BMP levels in brain tissue of APP/PSEN transgenic compared to age matched wild-type mice were also observed using this platform. This analytical methodology presented a high throughput LC-based approach incorporating simulated MS/MS database to identify and quantify BMP lipids as well as PG molecules.

A Bag of Wavelet Features for Snore Sound Classification.

Snore sound (SnS) classification can support a targeted surgical approach to sleep related breathing disorders. Using machine listening methods, we aim to find the location of obstruction and vibration within a subject's upper airway. Wavelet features have been demonstrated to be efficient in the recognition of SnSs in previous studies. In this work, we use a bag-of-audio-words approach to enhance the low-level wavelet features extracted from SnS data. A Naïve Bayes model was selected as the classifier based on its superiority in initial experiments. We use SnS data collected from 219 independent subjects under drug-induced sleep endoscopy performed at three medical centres. The unweighted average recall achieved by our proposed method is 69.4%, which significantly ([Formula: see text] one-tailed z-test) outperforms the official baseline (58.5%), and beats the winner (64.2%) of the INTERSPEECH COMPARE Challenge 2017 Snoring sub-challenge. In addition, the conventionally used features like formants, mel-scale frequency cepstral coefficients, subband energy ratios, spectral frequency features, and the features extracted by the OPENSMILE toolkit are compared with our proposed feature set. The experimental results demonstrate the effectiveness of the proposed method in SnS classification.

Quantitation of Lysosomal Trapping of Basic Lipophilic Compounds Using In Vitro Assays and In Silico Predictions Based on the Determination of the Full pH Profile of the Endo-/Lysosomal System in Rat Hepatocytes.

Lysosomal sequestration may affect the pharmacokinetics, efficacy, and safety of new basic lipophilic drug candidates potentially impacting their intracellular concentrations and tissue distribution. It may also be involved in drug-drug interactions, drug resistance, and phospholipidosis. However, currently there are no assays to evaluate the lysosomotropic behavior of compounds in a setting fully meeting the needs of drug discovery. We have, therefore, integrated a set of methods to reliably rank order, quantify, and calculate the extent of lysosomal sequestration in rat hepatocytes. An indirect fluorescence-based assay monitors the displacement of the fluorescence probe LysoTracker Red by test compounds. Using a lysosomal-specific evaluation algorithm allows one to generate IC50 values at lower than previously reported concentrations. The concentration range directly agrees with the concentration dependency of the lysosomal drug content itself directly quantified by liquid chromatography-tandem mass spectrometry and thus permits a quantitative link between the indirect and the direct trapping assay. Furthermore, we have determined the full pH profile and corresponding volume fractions of the endo-/lysosomal system in plated rat hepatocytes, enabling a more accurate in silico prediction of the extent of lysosomal trapping based only on pK a values as input, allowing early predictions even prior to chemical synthesis. The concentration dependency-i.e., the saturability of the trapping-can then be determined by the IC50 values generated in vitro. Thereby, a more quantitative assessment of the susceptibility of basic lipophilic compounds for lysosomal trapping is possible.

Snoring - An Acoustic Definition.

Objective- The distinction of snoring and loud breathing is often subjective and lies in the ear of the beholder. The aim of this study is to identify and assess acoustic features with a high suitability to distinguish these two classes of sound, in order to facilitate an objective definition of snoring based on acoustic parameters. Methods- A corpus of snore and breath sounds from 23 subjects has been used that were classified by 25 human raters. Using the openSMILE feature extractor, 6 373 acoustic features have been evaluated for their selectivity comparing SVM classification, logistic regression, and the recall of each single feature. Results- Most selective single features were several statistical functionals of the first and second mel frequency spectrum-generated perceptual linear predictive (PLP) cepstral coefficient with an unweighted average recall (UAR) of up to 93.8%. The best performing feature sets were low level descriptors (LLDs), derivatives and statistical functionals based on fast Fourier transformation (FFT), with a UAR of 93.0%, and on the summed mel frequency spectrum-generated PLP cepstral coefficients, with a UAR of 92.2% using SVM classification. Compared to SVM classification, logistic regression did not show considerable differences in classification performance. Conclusion- It could be shown that snoring and loud breathing can be distinguished by robust acoustic features. The findings might serve as a guidance to find a consensus for an objective definition of snoring compared to loud breathing.

Synchronization in Interpersonal Speech.

During both positive and negative dyadic exchanges, individuals will often unconsciously imitate their partner. A substantial amount of research has been made on this phenomenon, and such studies have shown that synchronization between communication partners can improve interpersonal relationships. Automatic computational approaches for recognizing synchrony are still in their infancy. In this study, we extend on previous work in which we applied a novel method utilizing hand-crafted low-level acoustic descriptors and autoencoders (AEs) to analyse synchrony in the speech domain. For this purpose, a database consisting of 394 in-the-wild speakers from six different cultures, is used. For each speaker in the dyadic exchange, two AEs are implemented. Post the training phase, the acoustic features for one of the speakers is tested using the AE trained on their dyadic partner. In this same way, we also explore the benefits that deep representations from audio may have, implementing the state-of-the-art Deep Spectrum toolkit. For all speakers at varied time-points during their interaction, the calculation of reconstruction error from the AE trained on their respective dyadic partner is made. The results obtained from this acoustic analysis are then compared with the linguistic experiments based on word counts and word embeddings generated by our word2vec approach. The results demonstrate that there is a degree of synchrony during all interactions. We also find that, this degree varies across the 6 cultures found in the investigated database. These findings are further substantiated through the use of 4,096 dimensional Deep Spectrum features.

Automated Classification of Airborne Pollen using Neural Networks.

Pollen allergies are considered as a global epidemic nowadays, as they influence more than a quarter of the worldwide population, with this percentage expected to rapidly increase because of ongoing climate change. To date, alerts on high-risk allergenic pollen exposure have been provided only via forecasting models and conventional monitoring methods that are laborious. The aim of this study is to develop and evaluate our own pollen classification model based on deep neural networks. Airborne allergenic pollen have been monitored in Augsburg, Bavaria, Germany, since 2015, using a novel automatic Bio-Aerosol Analyzer (BAA 500, Hund GmbH). The automatic classification system is compared and evaluated against our own, newly developed algorithm. Our model achieves an unweighted average precision of 83.0 % and an unweighted average recall of 77.1 % across 15 classes of pollen taxa. Automatic, real-time information on concentrations of airborne allergenic pollen will significantly contribute to the implementation of timely, personalized management of allergies in the future. It is already clear that new methods and sophisticated models have to be developed so as to successfully switch to novel operational pollen monitoring techniques serving the above need.

Deep Unsupervised Representation Learning for Abnormal Heart Sound Classification.

Given the world-wide prevalence of heart disease, the robust and automatic detection of abnormal heart sounds could have profound effects on patient care and outcomes. In this regard, a comparison of conventional and state-of-theart deep learning based computer audition paradigms for the audio classification task of normal, mild abnormalities, and moderate/severe abnormalities as present in phonocardiogram recordings, is presented herein. In particular, we explore the suitability of deep feature representations as learnt by sequence to sequence autoencoders based on the auDeep toolkit. Key results, gained on the new Heart Sounds Shenzhen corpus, indicate that a fused combination of deep unsupervised features is well suited to the three-way classification problem, achieving our highest unweighted average recall of 47.9% on the test partition.

Snoring classified: The Munich-Passau Snore Sound Corpus.

OBJECTIVE: Snoring can be excited in different locations within the upper airways during sleep. It was hypothesised that the excitation locations are correlated with distinct acoustic characteristics of the snoring noise. To verify this hypothesis, a database of snore sounds is developed, labelled with the location of sound excitation. METHODS: Video and audio recordings taken during drug induced sleep endoscopy (DISE) examinations from three medical centres have been semi-automatically screened for snore events, which subsequently have been classified by ENT experts into four classes based on the VOTE classification. The resulting dataset containing 828 snore events from 219 subjects has been split into Train, Development, and Test sets. An SVM classifier has been trained using low level descriptors (LLDs) related to energy, spectral features, mel frequency cepstral coefficients (MFCC), formants, voicing, harmonic-to-noise ratio (HNR), spectral harmonicity, pitch, and microprosodic features. RESULTS: An unweighted average recall (UAR) of 55.8% could be achieved using the full set of LLDs including formants. Best performing subset is the MFCC-related set of LLDs. A strong difference in performance could be observed between the permutations of train, development, and test partition, which may be caused by the relatively low number of subjects included in the smaller classes of the strongly unbalanced data set. CONCLUSION: A database of snoring sounds is presented which are classified according to their sound excitation location based on objective criteria and verifiable video material. With the database, it could be demonstrated that machine classifiers can distinguish different excitation location of snoring sounds in the upper airway based on acoustic parameters.

How does a slender tibia resist buckling? Effect of material, structural and geometric characteristics on buckling behaviour of the hindleg tibia in stick insect postembryonic development.

During the lifespan of the stick insect Carausius morosus, their long and narrow tibiae experience substantial compressive loads. The mechanical load on the tibiae increases as the weight of the insect rises. The increase in body weight is accompanied by a notable increase in the insect's body size and, accordingly, by an increase in the length of the tibiae. Both of these changes can raise the risk of buckling of the tibiae. In this study, we tracked changes in the material and geometric properties of the hindleg tibia of C. morosus during growth. The results show that although buckling (either by Euler buckling or local buckling) is the dominant failure mode under compression, the tibia is very capable of maintaining its buckling resistance in each postembryonic developmental stage. This is essentially the result of a compromise between the increasing slenderness of the tibia and its increasing material stiffness. The use of an optimal radius to thickness ratio, a soft resilin-dominated core, and chitin fibres oriented in both longitudinal and circumferential directions are presumably additional strategies preventing buckling of the tibia. This study, providing the first quantitative data on changes in the biomechanical properties of cuticle during the entire life of an insect, is expected to shed more light on the structure-property-function relationship in this complex biological composite.

Role of renal function in risk assessment of target non-attainment after standard dosing of meropenem in critically ill patients: a prospective observational study.

BACKGROUND: Severe bacterial infections remain a major challenge in intensive care units because of their high prevalence and mortality. Adequate antibiotic exposure has been associated with clinical success in critically ill patients. The objective of this study was to investigate the target attainment of standard meropenem dosing in a heterogeneous critically ill population, to quantify the impact of the full renal function spectrum on meropenem exposure and target attainment, and ultimately to translate the findings into a tool for practical application. METHODS: A prospective observational single-centre study was performed with critically ill patients with severe infections receiving standard dosing of meropenem. Serial blood samples were drawn over 4 study days to determine meropenem serum concentrations. Renal function was assessed by creatinine clearance according to the Cockcroft and Gault equation (CLCRCG). Variability in meropenem serum concentrations was quantified at the middle and end of each monitored dosing interval. The attainment of two pharmacokinetic/pharmacodynamic targets (100%T>MIC, 50%T>4×MIC) was evaluated for minimum inhibitory concentration (MIC) values of 2 mg/L and 8 mg/L and standard meropenem dosing (1000 mg, 30-minute infusion, every 8 h). Furthermore, we assessed the impact of CLCRCG on meropenem concentrations and target attainment and developed a tool for risk assessment of target non-attainment. RESULTS: Large inter- and intra-patient variability in meropenem concentrations was observed in the critically ill population (n = 48). Attainment of the target 100%T>MIC was merely 48.4% and 20.6%, given MIC values of 2 mg/L and 8 mg/L, respectively, and similar for the target 50%T>4×MIC. A hyperbolic relationship between CLCRCG (25-255 ml/minute) and meropenem serum concentrations at the end of the dosing interval (C8h) was derived. For infections with pathogens of MIC 2 mg/L, mild renal impairment up to augmented renal function was identified as a risk factor for target non-attainment (for MIC 8 mg/L, additionally, moderate renal impairment). CONCLUSIONS: The investigated standard meropenem dosing regimen appeared to result in insufficient meropenem exposure in a considerable fraction of critically ill patients. An easy- and free-to-use tool (the MeroRisk Calculator) for assessing the risk of target non-attainment for a given renal function and MIC value was developed. TRIAL REGISTRATION: Clinicaltrials.gov, NCT01793012 . Registered on 24 January 2013.