Search

To evaluate the accuracy of a new electronic nose to recognize prostate cancer in urine samples.A blind, prospective study on consecutive patients was designed. Overall, 174 subjects were included in the study: 88 (50.6%) in prostate cancer group, and 86 (49.4%) in control group. Electronic nose performance for prostate cancer was assessed using sensitivity and specificity. The diagnostic accuracy of electronic nose was reported as area under the receiver operating characteristic curve.The electronic nose in the study population reached a sensitivity 85.2% (95% confidence interval 76.1-91.9; 13 false negatives out of 88), a specificity 79.1% (95% confidence interval 69.0-87.1; 18 false positives out of 86). The accuracy of the electronic nose represented as area under the receiver operating characteristic curve 0.821 (95% confidence interval 0.764-0.879).The diagnostic accuracy of electronic nose for recognizing prostate cancer in urine samples is high, promising and susceptible to supplemental improvement. Additionally, further studies will be necessary to design a clinical trial to validate electronic nose application in diagnostic prostate cancer nomograms.

[ES] Encontrar un método no invasivo de clasificación fiable para distinguir entre pacientes con cancer de próstata e hiperplasia benigna de próstata no es una tarea trivial. El antígeno prostático específico mas utilizado presenta un au- mento en otras situaciones como la retención de orina o la eyaculación. Los productos metabólicos celulares como los compuestos volatiles en forma de bio-fluidos presentan una oportunidad para estudiarlos con la tecnología de semiconductores de óxido de metal. Una nariz electrónica diseñada para estos biofiuidos podría convertirse en una prueba ambulatoria clave para disminuir el número de biopsias. El diseño 3D, la creación de circuitos electrónicos, neumáticos, la adecuación de las muestras, la creación de una base de datos estandarizada para narices electrónicas, la definición de una interfaz de usuario amigable para el profesional sanitario, un sistema securizado para la comunicación con un servidor de almacenamiento de datos y todo el software necesario para hacer viable la democratización de las narices electrónicas en atención hospitalaria configuran la estrategia del trabajo. Además, el método empleado para conseguir aumentar el número de datos ha resultado determinante para hacer mas sencillo el entrenamiento de la red neuronal con la que se ha logrado una precisión del 87% en la clasificación entre pacientes de cancer de próstata e hiperplasia benigna de próstata. Se ha escogido un formato libro donde recoger todos los aspectos significativos de la realización del estudio. El primer capítulo se centra en el proyecto de tesis a financiar, la problemática a resolver, los objetivos y el plan de trabajo. En el segundo capítulo se definen los conceptos, el sentido del olfato, la vejiga, la próstata, los sensores utilizados, indicadores de cancer y otros elementos como la nariz electrónica. Es con una nariz electrónica de donde se obtienen las señales que se tratan en el capítulo tercero. En este capítulo se muestran las características de las señales provenientes de pacientes con cancer de próstata y pacientes con hiperplasia benigna de próstata, su tratamiento, filtrado, comparación y uso para la clasificación de pacientes. Asimismo, se muestran dos métodos distintos de clasificación y estudio de los algoritmos para automatizar la misma. El capítulo cuarto se centra en la construcción de un prototipo. Se ha abordado este capítulo como un manual de construcción que permita al lector conocer el proceso realizado, la toma de decisiones y la técnica utilizada para la realización del hardware, software y el modelo 3D. Dado que la diversidad de especialidades que confluyen en el mismo proyecto puede imposibilitar el dominio de alguno de los campos tratados, incluso dentro del mismo campo de especialización, se ha preparado este capítulo con diagramas y código fu- ente que acompaña a las explicaciones de forma detallada. El propósito de este capítulo es permitir a investigadores de diversas areas de conocimiento su construcción sin necesidad de profundizar en otros campos. Además, en este capítulo se detallan los procedimientos a fin de favorecer la democratización de la nariz electrónica como instrumento científico y con la esperanza de proporcionar la base para futuras implementaciones en la comunidad científica. Para el autor, la tesis no es sólo un documento, es un período de evolución personal. Ha estado condicionada por factores diversos, una pandemia, el planteamiento inicial, la financiación, el grupo de acogida, la línea de investigación, el desarrollo socio-económico del autor y el propio autoaprendizaje en el proceso de ejecución. Es por ello que se ha plasmado en el capítulo quinto lo que ha guiado de una u otra manera, pero de forma significativa, la toma de decisiones. Este capítulo versa sobre el trabajo futuro y tres proyectos, COST y H2020, DE1-SoC ADC e-Nose y Detección de tumores, que han ejercido influencia en el desarrollo de la tesis. Las conclusiones se recogen en el capítulo sexto. No obstante, cada capítulo dispone de un apartado llamado "Reflexiones" donde se anotan aquellas ideas y conclusiones concretas sobre el mismo. De esta forma se proporciona al lector un instrumento de lectura rápida que permita su consulta como si se tratara de un manual, sin necesidad de leer la totalidad del documento. En resumen, en este estudio se detallan los métodos y procedimientos de la investigación realizada con pacientes de cáncer de próstata e hiperplasia benigna de próstata con el propósito de conseguir un prototipo de una nariz electrónica para la clasificación de pacientes haciendo uso de orina como biofluido. Se ha tratado de escribir una referencia desde la que seguir construyendo conocimiento en torno a las narices electrónicas aplicadas a la clasificación de pacientes., [CA] Trobar una mètode no invasiu de classificació fiable per a distingir entre pacients amb càncer de pròstata i hiperplàsia benigna de pròstata no és una tasca trivial. L'antigen prostàtic específic més utilitzat presenta un augment en altres situacions com la retenció d'orina o l'ejaculació. Els productes metabòlics cel·lulars com els compostos volàtils en forma de bio-fluids presenten una oportunitat per tal d'estudiar-los amb la tecnologia de semiconductors d'òxid de metall. Un nas electrònic dissenyat per a aquestos bio-fluids podria convertir-se en una prova ambulatòria clau per tal de disminuir el nombre de biòpsies. El disseny 3D, la creació de circuits electrònics, pneumàtics, l'adequació de les mostres, la creació d'una base de dades estandarditzada per a nassos electrònics, la definició d'una interfície d'usuari amigable per al professional sanitari, un sistema securitzat per a la comunicació amb un servidor d'emmagatzemament de dades i tot el software necessari per tal de fer viable la democratització dels nassos electrònics en atenció hospitalària configuren l'estratègia del treball. A més, el mètode emprat per a aconseguir augmentar el nombre de dades ha resultat determinant per fer més senzill l'entrenament de la xarxa neuronal amb la qual s'ha aconseguit una precisió del 87% a la classificació entre pacients de càncer de pròstata i hiperplàsia benigna de pròstata. S'ha escollit un format llibre on recollir tots els indrets significatius de la realització de l'estudi. El primer capítol es centra en el projecte de tesi a finançar, la problemàtica a resoldre, els objectius i el pla de treball. Al segon capítol es defineixen els conceptes, el sentit de l'olfacte, la bufeta, la próstata, els sensors utilitzats, indicadors de càncer i altres elements com el nas electrònic. És amb un nas electrònic d'on s'obtenen les senyals que es tracten al capítol tercer. En aquest capítol es mostren les característiques dels senyals provinents de pacients amb càncer de pròstata i pacients amb hiperplàsia benigna de próstata, el seu tractament, filtrat, comparació i ús per a la classificació de pacients. Així mateix, es mostren dos mètodes diferents de classificació i estudi dels algorismes per a automatitzar la mateixa. El capítol quart es centra en la construcció d'un prototip. S'ha abordat aquest capítol com un manual de construcció que permeta al lector conéixer el procés realitzat, la presa de decisions i la tècnica emprada pera la realització tant del hardware, els software i com el model 3D. Com que la diversitat d'especialitats que conflueixen en el mateix projecte pot impossibilitar el domini d'algun dels camps tractats, inclòs dins del mateix camp d'especialització, s'ha preparat aquest capítol amb diagrames i codi font que acompanya les explicacions de forma detallada. El propòsit d'aquest capítol és permetre a investigadors de diverses àrees de coneixement la seva construcció sense necessitat d'aprofundir en altres camps. A més, en aquest capítol es detallen els procediments per tal d'afavorir la democratització del nas electrònic com a instrument científic i amb la esperança de proporcionar la base per a futures implementacions en la comunitat científica. Per a l'autor, la tesi no és sols un document, és un període d'evolució personal. Ha estat condicionada per factors diversos, una pandèmia, el plantejament inicial, el finançament, el grup d'acollida, la línia d'investigació, el desenvolupament socio-econòmic de l'autor i el propi autoaprenentatge al procés d'execució. És per això que s'ha plasmat al capítol cinqué allò que ha guiat d'una manera o d'altra, però de forma significativa, la presa de decisions. Aquest capítol versa sobre el treball futur i tres projectes, COST i H2020, DE1-SoC ADC e-Nose i Detecció de tumors, que han exercit influència al desenvolupament de la tesi. És un exercici de reflexió al voltant dels mateixos i són una mostra de la capacitat investigadora de l'autor. Les conclusions es recullen al capítol sisé. No obstant això, cada capítol disposa d'un apartat anomenat "Reflexions" on s'anoten aquelles idees i conclusions concretes sobre el mateix. D'aquesta manera es proporciona al lector un instrument de lectura ràpida que permeta la seva consulta com si es tractés d'un manual, sense necessitat de llegir la totalitat del document. En resum, en aquest estudi es detallen els mètodes i procediments de la investigació realitzada amb pacients de càncer de pròstata i hiperplàsia benigna de pròstata amb el propòsit d'aconseguir un prototip d'un nas electrònic per a la classificació de pacients fent ús d'orina com a bio-fluid. S'ha tractat d'escriure una referència des d'on seguir construint coneixement al voltant dels nassos electrònics aplicats a la classificació de pacients., [EN] Finding a reliable non-invasive classification method to distinguish between patients with prostate cancer and benign prostatic hyperplasia is not a trivial task. The most commonly used specific prostate antigen shows an increase in other situations such as urinary retention or ejaculation. Cellular metabolic products, such as volatile compounds in the form of bio- fiuids, present an opportunity to study them by using metal oxide semiconduc- tor technology. An electronic nose designed for these bio-fiuids could become a key ambulatory test in order to decrease the number of biopsies. The strategy of the work is configured in diferent fields. Design a 3D set to enclose the electronics and penumatic circuits. The suitability of the samples it has to be studied furthermore the creation of a standardized database for electronic noses to store these samples. The definition of a friendly user interface for the health professional besides a system secured for communication with a data storage server. To sum up, all the software and hardware necessary to make viable the democratization of electronic noses in hospital care. In addition, the method be inployed to increase the number of data has been decisive to simplify the training of the neural network with which an accuracy of 87% has been achieved in the classification between prostate cancer patients and benign prostatic hyperplasia. A book format has been chosen to collect all the significant points of the study. The first chapter focuses on the thesis project to be financed, the problem to be solved, the objectives and the work plan. The second chapter defines the concepts, the sense of smell, the bladder, the prostate, the sensors used, cancer indicators and other elements such as the electronic nose. It is with an electronic nose that the signals discussed in chapter three are obtained. This chapter shows the characteristics of signals from patients with prostate cancer and patients with benign prostatic hyper- plasia, their treatment, filtering, comparison and use for patient classification. Likewise, two different methods of classification and study of the algorithms to automate the same are shown. The fourth chapter focuses on the construction of a prototype. This chapter has been approached as a construction manual that allows the reader to know the process carried out, the decision-making and the technique used to make both the hardware, the software and the 3D. Since the diversity of specialties that converge in the same project can make it impossible to master any of the fields discussed, including within the same field of specialization, this chapter has been prepared with diagrams and source code that accompanies the explanations of form detailed. The purpose of this chapter is to allow researchers from various areas of knowledge to construct it without the need to delve into other fields. In addition, this chapter details the procedures to promote the democratization of the electronic nose as a scientific instrument and with the hope of providing the basis for future implementations in the scientific community. For the author, the thesis is not just a document, it is a period of personal evolution. It has been conditioned by various factors, a pandemic, the initial approach, the funding, the host group, the line of research, the socioof the author and the self-learning in the execution process. That is why what has been shaped in the fifth has guided in one way or another, but in a significant way, the decision-making. This chapter deals with future work and three projects, COST and H2020, DE1-SoC ADC eNose and Tumor Detection, which have influenced the development of the thesis. It is an exercise in reflection about them and they are a sample of the author's research capacity. The conclusions are collected in chapter six. However, each chapter has a section called "Reflections" where those specific ideas and conclusions about the same are noted.

BACKGROUND: Electronic noses (eNoses) are emerging point-of-care tools that may help in the subphenotyping of chronic respiratory diseases such as asthma.OBJECTIVE: We aimed to investigate whether eNoses can classify atopy in pediatric and adult patients with asthma.METHODS: Participants with asthma and/or wheezing from 4 independent cohorts were included; BreathCloud participants (n = 429), Unbiased Biomarkers in Prediction of Respiratory Disease Outcomes adults (n = 96), Unbiased Biomarkers in Prediction of Respiratory Disease Outcomes pediatric participants (n = 100), and Pharmacogenetics of Asthma Medication in Children: Medication with Anti-Inflammatory Effects 2 participants (n = 30). Atopy was defined as a positive skin prick test result (≥3 mm) and/or a positive specific IgE level (≥0.35 kU/L) for common allergens. Exhaled breath profiles were measured by using either an integrated eNose platform or the SpiroNose. Data were divided into 2 training and 2 validation sets according to the technology used. Supervised data analysis involved the use of 3 different machine learning algorithms to classify patients with atopic versus nonatopic asthma with reporting of areas under the receiver operating characteristic curves as a measure of model performance. In addition, an unsupervised approach was performed by using a bayesian network to reveal data-driven relationships between eNose volatile organic compound profiles and asthma characteristics.RESULTS: Breath profiles of 655 participants (n = 601 adults and school-aged children with asthma and 54 preschool children with wheezing [68.2% of whom were atopic]) were included in this study. Machine learning models utilizing volatile organic compound profiles discriminated between atopic and nonatopic participants with areas under the receiver operating characteristic curves of at least 0.84 and 0.72 in the training and validation sets, respectively. The unsupervised approach revealed that breath profiles classifying atopy are not confounded by other patient characteristics.CONCLUSION: eNoses accurately detect atopy in individuals with asthma and wheezing in cohorts with different age groups and could be used in asthma phenotyping.

Breath analysis using eNose technology can be used to discriminate between asthma and COPD patients, but it remains unclear whether results are influenced by smoking status. We aim to study whether eNose can discriminate between ever- vs. never-smokers and smoking &lt, 24 vs. &gt, 24 h before the exhaled breath, and if smoking can be considered a confounder that influences eNose results. We performed a cross-sectional analysis in adults with asthma or chronic obstructive pulmonary disease (COPD), and healthy controls. Ever-smokers were defined as patients with current or past smoking habits. eNose measurements were performed by using the SpiroNose. The principal component (PC) described the eNose signals, and linear discriminant analysis determined if PCs classified ever-smokers vs. never-smokers and smoking &lt, 24 h. The area under the receiver–operator characteristic curve (AUC) assessed the accuracy of the models. We selected 593 ever-smokers (167 smoked &lt, 24 h before measurement) and 303 never-smokers and measured the exhaled breath profiles of discriminated ever- and never-smokers (AUC: 0.74, 95% CI: 0.66–0.81), and no cigarette consumption &lt, 24h (AUC 0.54, 95% CI: 0.43–0.65). In healthy controls, the eNose did not discriminate between ever or never-smokers (AUC 0.54, 95% CI: 0.49–0.60) and recent cigarette consumption (AUC 0.60, 95% CI: 0.50–0.69). The eNose could distinguish between ever and never-smokers in asthma and COPD patients, but not recent smokers. Recent smoking is not a confounding factor of eNose breath profiles.

Exhaled breath analysis for early disease detection may provide a convenient method for painless and non-invasive diagnosis. In this work, a novel, compact and easy-to-use breath analyzer platform with a modular sensing chamber and direct breath sampling unit is presented. The developed analyzer system comprises a compact, low volume, temperature-controlled sensing chamber in three modules that can host any type of resistive gas sensor arrays. Furthermore, in this study three modular breath analyzers are explicitly tested for reproducibility in a real-life breath analysis experiment with several calibration transfer (CT) techniques using transfer samples from the experiment. The experiment consists of classifying breath samples from 15 subjects before and after eating a specific meal using three instruments. We investigate the possibility to transfer calibration models across instruments using transfer samples from the experiment under study, since representative samples of human breath at some conditions are difficult to simulate in a laboratory. For example, exhaled breath from subjects suffering from a disease for which the biomarkers are mostly unknown. Results show that many transfer samples of all the classes under study (in our case meal/no meal) are needed, although some CT methods present reasonably good results with only one class., publishedVersion

In the present work, a novel, portable and innovative eNose composed of a surface acoustic wave (SAW) sensor array based on zeolitic imidazolate frameworks, ZIF-8 and ZIF-67 nanocrystals (pure and combined with gold nanoparticles), as sensitive layers has been tested as a non-invasive system to detect different disease markers, such as acetone, ethanol and ammonia, related to the diagnosis and control of diabetes mellitus through exhaled breath. The sensors have been prepared by spin coating, achieving continuous sensitive layers at the surface of the SAW device. Low concentrations (5 ppm, 10 ppm and 25 ppm) of the marker analytes were measured, obtaining high sensitivities, good reproducibility, short time response and fast signal recovery., This work has been supported by Universidad Nacional Autónoma de México via Grants DGAPA-UNAM-PAPIIT TA100118 and DGAPA-UNAM-PAPIIT IT100518, the Fundación General CSIC via Programa ComFuturo, and the Spanish Ministry of Science and Innovation via Grant TEC2016-79898-C6-(AEI/FEDER,EU). This research has used the Spanish ICTS Network MICRONANOFABS (partially funded by MINECO), We acknowledge support by the CSIC Open Access Publication Initiative through its Unit of Information Resources for Research (URICI)

Fecal volatile organic compound (VOC) analysis has shown great potential as a noninvasive diagnostic biomarker for a variety of diseases. Before clinical implementation, the factors influencing the outcome of VOC analysis need to be assessed. Recent studies found that the sampling conditions can influence the outcome of VOC analysis. However, the dietary influences remains unknown, especially in (preterm) infants. Therefore, we assessed the effects of feeding composition on fecal VOC patterns of preterm infants (born at &lt, 30 weeks gestation). Two subgroups were defined: (1) daily intake &gt, 75% breastmilk (BM) feeding and (2) daily intake &gt, 75% formula milk (FM) feeding. Fecal samples, which were collected at 7, 14 and 21 days postnatally, were analyzed by an electronic nose device (Cyranose 320&reg, ). In total, 30 preterm infants were included (15 FM, 15 BM). No differences in the fecal VOC patterns were observed at the three predefined time-points. Combining the fecal VOC profiles of these time-points resulted in a statistically significant difference between the two subgroups although this discriminative accuracy was only modest (AUC [95% CI], p-value, sensitivity, and specificity of 0.64 [0.51&ndash, 0.77], 0.04, 68%, and 51%, respectively). Our results suggest that the influence of enteral feeding on the outcome of fecal VOC analysis cannot be ignored in this population. Furthermore, in both subgroups, the fecal VOC patterns showed a stable longitudinal course within the first month of life.

In recent years pollution due to hydrocarbon spills has increased markedly as a result of the numerous advances in technologies and industrial processes. Anthropogenic activities (accidental or illegal) are responsible for most of these incidents. In some cases, the spills are not detected at the moment they occur and the contaminants are subjected to different degradation phenomena that may change the chemical composition of the hydrocarbon over time. An incorrect or ineffective identification of the spill could lead to significant consequences, bearing in mind that most spills are hazardous to the environment. In the present work the capacity of the analytical technique based on the Electronic Nose (eNose) combined with chemometrics in the identification and discrimination of different weathered petroleum-derived products (PDPs) was studied. Different volumes (40 &mu, L and 80 &mu, L) of PDPs (gasoline, diesel, and paraffin) were poured onto different supports (wood, cork, paper, and cotton sheet) and subjected to a natural weathering process by evaporation for one month. The porosity of the support was also studied. The application of linear discriminant analysis allowed the full discrimination of the samples according to the presence/absence of PDP and a 97.7% of correct discrimination of the different PDPs regardless of the weathering time, support or volume used. The results show that the system is capable of detecting and discriminating the presence of petroleum-derived products in any of the situations studied.

Acute rhinosinusitis (ARS) is a sudden, symptomatic inflammation of the nasal and paranasal mucosa. It is usually caused by respiratory virus infection, but bacteria complicate for a small number of ARS patients. The differential diagnostics between viral and bacterial pathogens is difficult and currently no rapid methodology exists, so antibiotics are overprescribed. The electronic nose (eNose) has shown the ability to detect diseases from gas mixtures. Differential mobility spectrometry (DMS) is a next-generation device that can separate ions based on their different mobility in high and low electric fields. Five common rhinosinusitis bacteria (Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, Staphylococcus aureus, and Pseudomonas aeruginosa) were analysed in vitro with DMS. Classification was done using linear discriminant analysis (LDA) and k-nearest neighbour (KNN). The results were validated using leave-one-out cross-validation and separate train and test sets. With the latter, 77% of the bacteria were classified correctly with LDA. The comparative figure with KNN was 79%. In one train-test set, P. aeruginosa was excluded and the four most common ARS bacteria were analysed with LDA and KNN; the correct classification rate was 83 and 85%, respectively. DMS has shown its potential in detecting rhinosinusitis bacteria in vitro. The applicability of DMS needs to be studied with rhinosinusitis patients.

In a fire investigation the rapid detection of the presence of ignitable liquids like gasoline is of great importance as it allows appropriate treatment of the remains, the identification of prevention methods and detects the possible presence of an arsonist. In some cases, analysts cannot access the fire scene in the first few hours due to the dangers involved in the situation and, as a consequence, phenomena such as weathering start. Ignitable liquid weathering is an evaporation process that results in an increase in the abundance of non-volatile compounds relative to volatile compounds, and this process changes the chemical composition. In the present work, the weathering of samples of gasoline at different times (from 0 h to a month) has been studied using an electronic nose (eNose). The influence of the volume used (40 µL and 80 µL) and the type of support (cork, wood, paper and cotton sheet) has been studied. Chemometric tools have been used with the aim of ascertaining the weathering time for which the developed method is capable of detecting the presence of gasoline. The eNose was able to discriminate samples of weathered gasoline. The support used for the samples did not seem to have an influence on the detection and the system.

To obtain a better understanding of the complex clinical conditions and disease mechanisms of severe asthma, the Unbiased Biomarkers for the Prediction of Respiratory Disease Outcomes (U-BIOPRED) consortium was initiated. By generating multi-molecular patient profiles and revealing distinct phenotypes, the consortium aims for more accurate prediction of therapeutic efficacy, and identification of novel targets and mechanistic pathways in both adult and pediatric patients. Noninvasive metabolomics of exhaled air (breathomics) through gas-sensor driven electronic nose (eNose) technology and gas chromatography–mass spectrometry (GC-MS) were two of the many applied omics approaches during this project. Challenging study aspects such as the incorporation of multiple eNose devices and a pan-European multicenter setting led to the establishment of the U-BIOPRED eNose platform, and the development of new sampling and storage strategies of exhaled breath samples. Based on a description of adapted technologies, and by virtue of a proof of concept analysis, this chapter provides a first glimpse into the outcomes of the U-BIOPRED eNose and GC-TOF-MS datasets, combined with information about new insights and future directions for the field of breathomics.

Electronic nose (eNose) systems are particularly appreciated for their portability, usability, relative low cost, and real-time or near real-time response. Their application finds space in several domains, including environmental monitoring. Within this field, marine monitoring is of particular scientific relevance due to the fragility of this specific environment, daily threatened by human activities that can potentially bring to catastrophic and irreversible consequences on marine wildlife. Under such considerations, a systematic review, complying with the PRISMA guidelines, was conducted covering the period up to 15 October 2018, in PubMed, ScienceDirect, and Google Scholar. Despite the relatively low number of articles published on this specific topic and the heterogeneity of the technological approaches employed, the results obtained by the various groups highlight the positive contribution eNose has given and can provide in near future for the monitoring and safeguarding of this delicate environment.

Respiratory infections, viral or bacterial, are a common cause of acute exacerbations of chronic obstructive pulmonary disease (AECOPD). A rapid, point-of-care, and easy-to-use tool distinguishing viral and bacterial from other causes would be valuable in routine clinical care. An electronic nose (e-nose) could fit this profile but has never been tested in this setting before. In a single-center registered trial (NTR 4601) patients admitted with AECOPD were tested with the Aeonose (R) electronic nose, and a diagnosis of viral or bacterial infection was obtained by bacterial culture on sputa and viral PCR on nose swabs. A neural network with leave-10%-out cross-validation was used to assess the e-nose data. Forty three patients were included. In the bacterial infection model, 22 positive cases were tested versus the negatives; and similarly 18 positive cases were tested in the viral infection model. The Aeonose was able to distinguish between COPD-subjects suffering from a viral infection and COPD patients without infection, showing an area under the curve (AUC) of 0.74. Similarly, for bacterial infections, an AUC of 0.72 was obtained. The Aeonose e-nose yields promising results in 'smelling' the presence or absence of a viral or bacterial respiratory infection during an acute exacerbation of COPD. Validation of these results using a new and large cohort is required before introduction into clinical practice.

Currently, many different methods are being used for pre-processing, statistical analysis and validation of data obtained by electronic nose technology from exhaled air. These various methods, however, have never been thoroughly compared. We aimed to empirically evaluate and compare the influence of different dimension reduction, classification and validation methods found in published studies on the diagnostic performance in several datasets. Our objective was to facilitate the selection of appropriate statistical methods and to support reviewers in this research area. We reviewed the literature by searching Pubmed up to the end of 2014 for all human studies using an electronic nose and methodological quality was assessed using the QUADAS-2 tool tailored to our review. Forty-six studies were evaluated regarding the range of different approaches to dimension reduction, classification and validation. From forty-six reviewed articles only seven applied external validation in an independent dataset, mostly with a case-control design. We asked their authors to share the original datasets with us. Four of the seven datasets were available for re-analysis. Published statistical methods for eNose signal analysis found in the literature review were applied to the training set of each dataset. The performance (area under the receiver operating characteristics curve (ROC-AUC)) was calculated for the training cohort (in-set) and after internal validation (leave-one-out cross validation). The methods were also applied to the external validation set to assess the external validity of the performance. Risk of bias was high in most studies due to non-random selection of patients. Internal validation resulted in a decrease in ROC-AUCs compared to in-set performance: -0.15,-0.14,-0.1,-0.11 in dataset 1 through 4, respectively. External validation resulted in lower ROC-AUC compared to internal validation in dataset 1 (-0.23) and 3 (-0.09). ROC-AUCs did not decrease in dataset 2 (+0.07) and 4 (+0.04). No single combination of dimension reduction and classification methods gave consistent results between internal and external validation sets in this sample of four datasets. This empirical evaluation showed that it is not meaningful to estimate the diagnostic performance on a training set alone, even after internal validation. Therefore, we recommend the inclusion of an external validation set in all future eNose projects in medicine.