Τα τελευταία χρόνια τόσο σε διεθνές όσο και σε εθνικό επίπεδο έχει δοθεί βαρύτητα στην ανίχνευση της εμβρυϊκής υποξίας κατά τη διάρκεια της κύησης και του τοκετού, καθώς είναι σαφής η άμεση συσχέτισή της με βραχυπρόθεσμες και απώτερες επιπλοκές του νεογνού. Η επιτυχής έκβαση ενός τοκετού εξαρτάται κυρίως από τον έγκαιρο εντοπισμό της δημιουργίας της εμβρυϊκής υποξίας και η ανάγκη για την υλοποίηση υπολογιστικών συστημάτων για την έγκαιρη διάγνωσή της είναι συνεχής και αυξανόμενη. Στα πλαίσια της διδακτορικής διατριβής σχεδιάστηκε και αναπτύχθηκε ένα πρωτότυπο σύστημα διάγνωσης της πρόωρης οξέωσης του εμβρύου κατά τη διάρκεια του τοκετού, το οποίο βασίστηκε στην επεξεργασία του εμβρυϊκού καρδιακού ρυθμού και στην ανάλυση της εμβρυϊκής παλμικής οξυμετρίας. Στόχος του συγκεκριμένου συστήματος είναι να αποτελέσει ένα βοηθητικό σύστημα διάγνωσης της εμβρυϊκής υποξίας και να συμβάλει στη λήψη αποφάσεων σχετικά με το χρόνο αποπεράτωσης του τοκετού, με σκοπό την πρόληψη άμεσων και απώτερων σοβαρών νεογνικών επιπλοκών. Ειδικότερα, η υλοποίηση του προτεινόμενου συστήματος βασίστηκε στην εφαρμογή του συνεχούς μετασχηματισμού κυματιδίων και της προσαρμοστικής προσέγγισης με τη χρήση του αλγορίθμου matching pursuit για την ανάδειξη της «κρυμμένης» πληροφορίας που μεταφέρει ο εμβρυϊκός καρδιακός ρυθμός στις πολύ χαμηλές συχνότητες. Συνδυάζοντας τα αποτελέσματα των παραπάνω τεχνικών επεξεργασίας, και ιδιαίτερα του αλγορίθμου matching pursuit, με τον προσδιορισμό του συνολικού χρόνου χαμηλού κορεσμού του αρτηριακού εμβρυϊκού αίματος, υλοποιήθηκε, με τη χρήση ενός διαμεριστικού αλγορίθμου, το προτεινόμενο σύστημα. Η ανάπτυξη αυτού του συστήματος αποτελεί μια καινοτόμα και πολλά υποσχόμενη προσέγγιση στην προσπάθεια της διάγνωσης της εμβρυϊκής υποξίας δεδομένου ότι παρουσιάζει υψηλή ειδικότητα και θετική προγνωστική αξία συμβάλοντας στην επίλυση του σοβαρότερου μειονεκτήματος της κλασσικής καρδιοτοκογραφίας που είναι η χαμηλή τιμή των αντίστοιχων προγνωστικών δεικτών. / In the last few years the research community has given great attention to the detection of antepartum and intrapartum fetal hypoxia, given its direct impact on both short- and long-term neonatal morbidity and mortality. The successful completion of labor depends mainly on the prompt identification of fetal hypoxia. The development of computational systems for the early diagnosis of restricted fetal oxygen supply is therefore of critical importance. This thesis involves the design and development of an innovative system for the early detection of acidosis, which was based on the fetal heart rate processing and fetal pulse oximetry analysis. The aim of the system is to comprise a computer-aided diagnostic system of fetal hypoxia and to contribute to the decision making regarding the labor completion time, in order to prevent short- and long-term neonatal complications. Specifically, the development of the system was based on the implementation of continuous wavelet transform and adaptive approximation using the matching pursuit algorithm, in order to reveal the “hidden” information conveyed in the very low frequency range of the fetal heart rate. The system involves the combination of the results of the above-mentioned processing techniques, and especially of the matching pursuit algorithm, along with the calculation of the duration of fetal arterial low oxygen saturation, applying a commonly used clustering algorithm. The proposed system is an innovative and promising approach towards the early diagnosis of fetal hypoxia, given its high specificity and positive predictive value, thus effectively addressing the major drawback of clinical cardiotocography.

O objetivo do presente estudo foi verificar a frequência cardíaca fetal e as respostas cardiorrespiratórias de gestantes e não-gestantes, durante e ao longo de 30 minutos após a execução de exercícios de força para membros superiores e inferiores, em dois volumes distintos. A amostra desse estudo foi composta por 20 mulheres, com idade entre 20 e 32 anos, sendo 10 gestantes (com idade gestacional entre 22 e 24 semanas) e 10 não-gestantes, que realizaram cinco sessões experimentais: sessão 1: familiarização com os equipamentos de coletas de dados e determinação de uma repetição máxima estimada; sessões 2, 3, 4 e 5: coleta das variáveis cardiorrespiratórias durante e ao longo de 30 minutos após os exercícios de força nos equipamentos cadeira extensora de joelhos bilateral e voador peitoral, com 1 e 3 séries de 15 repetições, com carga de 50% de uma repetição máxima estimada. Utilizou-se ANOVA para medidas repetidas com 2 e 3 fatores, com post-hoc de Bonferroni (α=0,05) (SPSS vs 13.0). Os resultados demonstraram que as respostas de pressão arterial durante os exercícios de força apresentam um comportamento mais baixo no grupo gestantes. Quando estes foram realizados com série única, somente a frequência cardíaca e o duplo produto apresentaram respostas diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral, entretanto, quando os exercícios foram realizados com séries múltiplas, as respostas das variáveis frequência cardíaca, pressão arterial sistólica, diastólica e média, duplo produto, ventilação e consumo de oxigênio absoluto foram diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral. As variáveis analisadas apresentaram diferenças ao longo dos 30 minutos de recuperação pós-exercício, em geral, retornando aos valores basais após 10 minutos do término do exercício. Não houve ocorrência de contrações uterinas em nenhuma gestante ao longo deste período. A resposta da frequência cardíaca fetal não apresentou diferenças nos diferentes exercícios e volumes e ao longo dos 30 minutos de recuperação pós-exercício e se manteve dentro dos padrões de normalidade (120-160 bpm). Conclui-se que, durante a realização dos exercícios de força extensor de joelhos bilateral e voador peitoral, as respostas de pressão arterial de gestantes foram menores do que as não-gestantes, o exercício extensor de joelhos bilateral apresentou maiores valores das variáveis cardiorrespiratórias comparado ao exercício voador e as variáveis apresentaram aumento das suas respostas com o aumento do número de séries realizadas. A resposta fetal não diferiu entre os exercícios e volumes. / The aim of the present study was to verify fetal heart rate and the cardiorespiratory responses in pregnant and non-pregnant women during and along 30 minutes postexecution of resistance exercises for upper and lower body, with two different volumes. The sample was composed of 20 healthy women, aged between 20-32 years old, being 10 pregnant women (gestational age between 22 and 24 weeks) and 10 non-pregnant women, who performed five experimental sessions: session 1: familiarization with the equipments and the determination of one estimated maximum repetition (1-RM); sessions 2, 3, 4 and 5: determination of the cardiorespiratory responses during and along 30 minutes post-resistance exercise on the bilateral leg extension and fly, with 1 and 3 sets of 15 repetitions, 50% of 1-RM. Results were analyzed using ANOVA for repeated measures with two and three factors with Bonferroni correction for post-hoc comparisons (α=0.05) (SPSS vs 13.0). The blood pressure responses during resistance exercises showed a lower behavior in the pregnant group. When the exercises were performed with a single set, only heart rate and rate-pressure product showed different responses between exercises, with increased values for bilateral leg extension, however, when the exercises were performed with multiple sets the heart rate, systolic, diastolic and mean blood pressure, rate-pressure product, ventilation and oxygen uptake responses were different between exercises, with increased values for bilateral leg extension. The analyzed variables showed differences along 30 minutes post-exercise but in general it was similar to those pre-exercise values after 10 minutes from the end of the resistance exercise. There was no occurrence of uterine contractions along this period. The fetal heart rate responses did not presented differences between the exercises and volumes and along the 30 minutes post-resistance exercise and kept the normality patterns (120-160 bpm). In conclusion, during the performance of the bilateral leg extension and fly resistance exercises the blood pressure response was lower in the pregnant group. The bilateral leg extension showed higher values for the cardiorespiratory variables compared to fly exercise and the variables responses presented an increase with the addition of the sets performed. The fetal response was not different between exercises and volumes performed.

O objetivo do presente estudo foi verificar a frequência cardíaca fetal e as respostas cardiorrespiratórias de gestantes e não-gestantes, durante e ao longo de 30 minutos após a execução de exercícios de força para membros superiores e inferiores, em dois volumes distintos. A amostra desse estudo foi composta por 20 mulheres, com idade entre 20 e 32 anos, sendo 10 gestantes (com idade gestacional entre 22 e 24 semanas) e 10 não-gestantes, que realizaram cinco sessões experimentais: sessão 1: familiarização com os equipamentos de coletas de dados e determinação de uma repetição máxima estimada; sessões 2, 3, 4 e 5: coleta das variáveis cardiorrespiratórias durante e ao longo de 30 minutos após os exercícios de força nos equipamentos cadeira extensora de joelhos bilateral e voador peitoral, com 1 e 3 séries de 15 repetições, com carga de 50% de uma repetição máxima estimada. Utilizou-se ANOVA para medidas repetidas com 2 e 3 fatores, com post-hoc de Bonferroni (α=0,05) (SPSS vs 13.0). Os resultados demonstraram que as respostas de pressão arterial durante os exercícios de força apresentam um comportamento mais baixo no grupo gestantes. Quando estes foram realizados com série única, somente a frequência cardíaca e o duplo produto apresentaram respostas diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral, entretanto, quando os exercícios foram realizados com séries múltiplas, as respostas das variáveis frequência cardíaca, pressão arterial sistólica, diastólica e média, duplo produto, ventilação e consumo de oxigênio absoluto foram diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral. As variáveis analisadas apresentaram diferenças ao longo dos 30 minutos de recuperação pós-exercício, em geral, retornando aos valores basais após 10 minutos do término do exercício. Não houve ocorrência de contrações uterinas em nenhuma gestante ao longo deste período. A resposta da frequência cardíaca fetal não apresentou diferenças nos diferentes exercícios e volumes e ao longo dos 30 minutos de recuperação pós-exercício e se manteve dentro dos padrões de normalidade (120-160 bpm). Conclui-se que, durante a realização dos exercícios de força extensor de joelhos bilateral e voador peitoral, as respostas de pressão arterial de gestantes foram menores do que as não-gestantes, o exercício extensor de joelhos bilateral apresentou maiores valores das variáveis cardiorrespiratórias comparado ao exercício voador e as variáveis apresentaram aumento das suas respostas com o aumento do número de séries realizadas. A resposta fetal não diferiu entre os exercícios e volumes. / The aim of the present study was to verify fetal heart rate and the cardiorespiratory responses in pregnant and non-pregnant women during and along 30 minutes postexecution of resistance exercises for upper and lower body, with two different volumes. The sample was composed of 20 healthy women, aged between 20-32 years old, being 10 pregnant women (gestational age between 22 and 24 weeks) and 10 non-pregnant women, who performed five experimental sessions: session 1: familiarization with the equipments and the determination of one estimated maximum repetition (1-RM); sessions 2, 3, 4 and 5: determination of the cardiorespiratory responses during and along 30 minutes post-resistance exercise on the bilateral leg extension and fly, with 1 and 3 sets of 15 repetitions, 50% of 1-RM. Results were analyzed using ANOVA for repeated measures with two and three factors with Bonferroni correction for post-hoc comparisons (α=0.05) (SPSS vs 13.0). The blood pressure responses during resistance exercises showed a lower behavior in the pregnant group. When the exercises were performed with a single set, only heart rate and rate-pressure product showed different responses between exercises, with increased values for bilateral leg extension, however, when the exercises were performed with multiple sets the heart rate, systolic, diastolic and mean blood pressure, rate-pressure product, ventilation and oxygen uptake responses were different between exercises, with increased values for bilateral leg extension. The analyzed variables showed differences along 30 minutes post-exercise but in general it was similar to those pre-exercise values after 10 minutes from the end of the resistance exercise. There was no occurrence of uterine contractions along this period. The fetal heart rate responses did not presented differences between the exercises and volumes and along the 30 minutes post-resistance exercise and kept the normality patterns (120-160 bpm). In conclusion, during the performance of the bilateral leg extension and fly resistance exercises the blood pressure response was lower in the pregnant group. The bilateral leg extension showed higher values for the cardiorespiratory variables compared to fly exercise and the variables responses presented an increase with the addition of the sets performed. The fetal response was not different between exercises and volumes performed.

O objetivo do presente estudo foi verificar a frequência cardíaca fetal e as respostas cardiorrespiratórias de gestantes e não-gestantes, durante e ao longo de 30 minutos após a execução de exercícios de força para membros superiores e inferiores, em dois volumes distintos. A amostra desse estudo foi composta por 20 mulheres, com idade entre 20 e 32 anos, sendo 10 gestantes (com idade gestacional entre 22 e 24 semanas) e 10 não-gestantes, que realizaram cinco sessões experimentais: sessão 1: familiarização com os equipamentos de coletas de dados e determinação de uma repetição máxima estimada; sessões 2, 3, 4 e 5: coleta das variáveis cardiorrespiratórias durante e ao longo de 30 minutos após os exercícios de força nos equipamentos cadeira extensora de joelhos bilateral e voador peitoral, com 1 e 3 séries de 15 repetições, com carga de 50% de uma repetição máxima estimada. Utilizou-se ANOVA para medidas repetidas com 2 e 3 fatores, com post-hoc de Bonferroni (α=0,05) (SPSS vs 13.0). Os resultados demonstraram que as respostas de pressão arterial durante os exercícios de força apresentam um comportamento mais baixo no grupo gestantes. Quando estes foram realizados com série única, somente a frequência cardíaca e o duplo produto apresentaram respostas diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral, entretanto, quando os exercícios foram realizados com séries múltiplas, as respostas das variáveis frequência cardíaca, pressão arterial sistólica, diastólica e média, duplo produto, ventilação e consumo de oxigênio absoluto foram diferentes entre os exercícios, com maiores valores no exercício extensor de joelhos bilateral. As variáveis analisadas apresentaram diferenças ao longo dos 30 minutos de recuperação pós-exercício, em geral, retornando aos valores basais após 10 minutos do término do exercício. Não houve ocorrência de contrações uterinas em nenhuma gestante ao longo deste período. A resposta da frequência cardíaca fetal não apresentou diferenças nos diferentes exercícios e volumes e ao longo dos 30 minutos de recuperação pós-exercício e se manteve dentro dos padrões de normalidade (120-160 bpm). Conclui-se que, durante a realização dos exercícios de força extensor de joelhos bilateral e voador peitoral, as respostas de pressão arterial de gestantes foram menores do que as não-gestantes, o exercício extensor de joelhos bilateral apresentou maiores valores das variáveis cardiorrespiratórias comparado ao exercício voador e as variáveis apresentaram aumento das suas respostas com o aumento do número de séries realizadas. A resposta fetal não diferiu entre os exercícios e volumes. / The aim of the present study was to verify fetal heart rate and the cardiorespiratory responses in pregnant and non-pregnant women during and along 30 minutes postexecution of resistance exercises for upper and lower body, with two different volumes. The sample was composed of 20 healthy women, aged between 20-32 years old, being 10 pregnant women (gestational age between 22 and 24 weeks) and 10 non-pregnant women, who performed five experimental sessions: session 1: familiarization with the equipments and the determination of one estimated maximum repetition (1-RM); sessions 2, 3, 4 and 5: determination of the cardiorespiratory responses during and along 30 minutes post-resistance exercise on the bilateral leg extension and fly, with 1 and 3 sets of 15 repetitions, 50% of 1-RM. Results were analyzed using ANOVA for repeated measures with two and three factors with Bonferroni correction for post-hoc comparisons (α=0.05) (SPSS vs 13.0). The blood pressure responses during resistance exercises showed a lower behavior in the pregnant group. When the exercises were performed with a single set, only heart rate and rate-pressure product showed different responses between exercises, with increased values for bilateral leg extension, however, when the exercises were performed with multiple sets the heart rate, systolic, diastolic and mean blood pressure, rate-pressure product, ventilation and oxygen uptake responses were different between exercises, with increased values for bilateral leg extension. The analyzed variables showed differences along 30 minutes post-exercise but in general it was similar to those pre-exercise values after 10 minutes from the end of the resistance exercise. There was no occurrence of uterine contractions along this period. The fetal heart rate responses did not presented differences between the exercises and volumes and along the 30 minutes post-resistance exercise and kept the normality patterns (120-160 bpm). In conclusion, during the performance of the bilateral leg extension and fly resistance exercises the blood pressure response was lower in the pregnant group. The bilateral leg extension showed higher values for the cardiorespiratory variables compared to fly exercise and the variables responses presented an increase with the addition of the sets performed. The fetal response was not different between exercises and volumes performed.

La surveillance du bien-être fœtal pendant le travail repose essentiellement sur l’enregistrement du rythme cardiaque fœtal (RCF). Celui-ci, même continu pendant le travail, ne permet pas d’évaluer parfaitement l’oxygénation du fœtus ni le risque d’asphyxie néonatale. En effet, cet outil est imparfait et son évaluation subjective avec une importante variabilité d’interprétation inter et intra opérateur. Des examens dits de seconde ligne sont utilisés en pratique courante pour caractériser l’état fœtal : le prélèvement de sang fœtal au scalp pour l’étude de l’équilibre acido-basique du fœtus (pH ou lactates) ou la pose d’électrode au scalp pour étudier l’ECG fœtal (analyse du segment ST). Ces techniques sont néanmoins invasives et sont soumises à des contraintes techniques. Il y a donc un intérêt à développer des moyens d’évaluation du bien être fœtal à la fois objectifs et non invasifs afin de diminuer la survenue d’une asphyxie périnatale. En effet, celle-ci touche 3 à 8 nouveaux nés pour 1000 naissances. La mortalité en période post-natale est de 25 à 50% des cas et ceux qui survivent développeront des troubles sévères (épilepsie, retard neuro-cognitif et comportemental, paralysie cérébrale…). Au cours de l’accouchement, l’asphyxie périnatale se caractérise par une diminution du pH artériel ombilical. Cette mesure du pH sanguin est donc la mesure de référence pour déterminer la sévérité de l’asphyxie.Une des voies étudiées pour améliorer le dépistage des fœtus à risque d’acidose est l’analyse des modifications du système nerveux autonome (SNA) par analyse de la variabilité du rythme cardiaque fœtal. En effet, la fréquence cardiaque fœtale est en permanence sous l’influence du système nerveux autonome et sa variabilité (VFC) est un reflet de la balance sympathique / parasympathique. Le CHU de Lille a développé une nouvelle méthode d’analyse continue de la VFC ayant montré son efficacité chez l’adulte et chez le nouveau né pour l’évaluation du SNA. L’objectif de ce travail de Thèse est d’adapter cette technologie à l’analyse du SNA fœtal pour obtenir un nouvel indice appelé Fetal Stress Index (FSI) et d’évaluer sa pertinence en situation d’acidose.Cette preuve de concept a été effectuée de manière expérimentale chez le fœtus de brebis. Elle s’est réalisée en 2 temps. Nous avons tout d’abord évalué la performance du FSI par rapport aux méthodes classiques d’analyse de la VFC en termes d’aptitude à détecter les variations du SNA. Après injection d’Atropine, parasympatholytique, ou de Propranolol, sympatholytique, nous avons montré que le FSI était une méthode efficace et spécifique d’évaluation des variations du tonus parasympathique du SNA. Cette étude a également montrée la supériorité du FSI par rapport aux méthodes classiques d’analyse de la VFC en termes de sensibilité et de spécificité. Dans un second temps, nous avons évalué ce nouvel indice comme facteur prédictif de l’état acido basique du fœtus dans 2 modèles expérimentaux d’occlusion cordonale. Dans le premier modèle, l’acidose était obtenue par une occlusion continue du cordon avec une réduction de partielle du débit ombilical. Dans le second, nous réalisions des occlusions totales répétées à intervalles réguliers afin de mimer les contractions utérines lors du travail. Dans les 2 études, nous avons observé une hausse du FSI en cas d’acidose avec une corrélation significative entre le FSI et le pH, mais aussi entre le FSI et les lactates dans le second modèle.En conclusion, le FSI constitue un bon reflet de l’activité parasympathique fœtale. Cet indice permet d’étudier les variations du SNA fœtal avec une meilleure sensibilité et une meilleure spécificité que les méthodes usuelles d’analyse de la VFC et semble bien corrélé à l’état acido basique fœtal. Il s’agit donc d’un indice prometteur qu’il sera intéressant d’incorporer dans une analyse multi paramétrique du rythme cardiaque fœtal. / The monitoring of fetal well being during labor is essentially based on fetal heart rate (FHR) analysis. The recording of FHR, even continuously during labor, does not fully assess fetal oxygenation or neonatal risk of asphyxia. Indeed, this tool is imperfect and subjective with an important inter and intra-operator variability. Second-line examinations to characterize the fetal state are currently used in routine practice, i.e. scalp fetal blood sampling to study the fetal acid-base balance (pH or lactates) or scalp electrode placement to study the fetal ECG (ST segment analysis). These techniques are nevertheless invasive and subject to technical constraints. There is therefore an interest in developing both objective and non-invasive means of evaluating fetal wellbeing to reduce neonatal encephalopathy. Indeed, its prevalence is about 3 to 8 per 1000 births. Post natal mortality is about 25 to 50% and survivors will hav severe diseases (epilepsy, neurologic impairment, cerebral palsy…).One of the possibilities studied to better identify fetuses at risk for acidosis is the analysis of changes in the autonomic nervous system (ANS) in response to hypoxia.Indeed, the regulation of heart rate is dependent on the ANS and thus, its variability is a reflection of the sympathetic / parasympathetic balance. Analysis of heart rate variability (HRV) is a recognized non-invasive tool that is used to assess ANS regulation. The CHU Lille has developed a new continuous tool for the analysis of HRV, which demonstrated its efficacity in adults and neonates to evaluate the ANS. The objective of this thesis was to develop its index, called Fetal Stress Index (FSI), in the fetus and to evaluate it in conditions of acidosis.The study was experimental in a sheep model chronically instrumented and was in 2 steps. First, we evaluate the performance of our method compared to commonly used HRV analysis, regarding the ability to detect the variation of variations of the ANS. After injection of atropine, to inhibit parasympathetic tone, or propranolol to block sympathetic activity, we shown that our method appeared to be effective in detecting parasympathetic inhibition and, moreover, was superior to classical analysis of HRV in terms of sensibility and specificity.In a second time, we evaluated this new index as a predictive factor of the fetal acid-base state in 2 experimental models of fetal hypoxia by occlusion of the cord. In the first one, acidosis was obtained through a partial occlusion of the umbilical cord and in the second one, though repetitive complete occlusion as uterine contractions during labor. In those two studies, we observed a raise of our index in case of acidosis with a correlation beetween FSI and pH and also FSI and lactates in the second model.In conclusion, the FSI reflects fetal parasympathetic activity, has a better detection than others usual methods, and seems well correlated to fetal acid-base status. It is a promising index and it will be interesting to incorporate it in a multi parametric analysis of fetal heart rate to predict acidosis.

Die nicht-invasive Analyse der fetalen Herzratenvariabilität mittels abdominaler Elektrokardiographie stellt eine neue Methode zur Beurteilung des fetalen Zustandes dar. Die Herzratenvariabilität gilt hier als ein sensitives Maß der autonomen Regulation. Bereits mit Beginn der zweiten Schwangerschaftshälfte ist es möglich, über das mütterliche Abdomen ein fetales Elektrokardiogramm abzuleiten und einer Herzratenvariabilitätsanalyse zuzuführen. Das Untersuchungskollektiv dieser Arbeit umfasste Frauen mit frühem vorzeitigem Blasensprung als Modell einer pathologischen Alteration des fetalen Zustandes sowie Frauen mit normalen Schwangerschaften zwischen der 20. und 28. Schwangerschaftswoche. Die technische Umsetzung und Analyse der fetalen Herzratenvariabilität erfolgte in Kooperation mit dem Institut für Biomedizinische Technik der TU Dresden. Insgesamt wurden 25 Datensätze der Auswertung zugeführt. Eine Reifung des autonomen Nervensystems des Feten mit ansteigendem Gestationsalter konnte mittels der Herzratenvariabilitätsanalyse gezeigt werden. Zur Risikostratifizierung der Feten bei frühem vorzeitigem Blasensprung konnten bei der Betrachtung der Herzratenvariabilitätsparameter keine signifikanten Unterschiede zum Normalkollektiv erhoben werden. Die Analyse der Parameter erfolgte hierbei unabhängig von den fetalen Verhaltenszuständen. Es konnte gezeigt werden, dass die nicht-invasive Analyse der Herzratenvariabilitätsanalyse methodisch und technisch in der Lage ist, den Fetalzustand und dessen Alterationen zu erfassen.

Despite the unprecedented improvement in child health in last 15 years, burden of stillbirth and neonatal death remain the key challenge in Nepal and the reduction of these deaths will be crucial for reaching the health targets for Sustainable development goal by 2030. The aim of this thesis was to explore the risk factors for stillbirth and neonatal death and change in perinatal outcomes after the introduction of the Helping Babies Breathe Quality Improvement Cycle (HBB QIC) in Nepal. This was a prospective cohort study with a nested case-control design completed in a tertiary hospital in Nepal. Information were collected from the women who had experienced perinatal death and live birth among referent population; a video recording was done in the neonatal resuscitation corner to collect information on the health workers’ performance in neonatal resuscitation.  Lack of antenatal care had the highest association with antepartum stillbirth (aOR 4.2, 95% CI 3.2–5.4), births that had inadequate fetal heart rate monitoring were associated with intrapartum stillbirth (aOR 1.9, CI 95% 1.5–2.4), and babies who were born premature and small-for-gestational-age had the highest risk for neonatal death in the hospital (aOR 16.2, 95% CI 12.3–21.3). Before the introduction of the HBB QIC, health workers displayed poor adherence to the neonatal resuscitation protocol. After the introduction of HBB QIC, the health workers demonstrated improvement in their neonatal resuscitation skills and these were retained until six months after training. Daily bag-and-mask skill checks (RR 5.1 95% CI 1.9–13.5), preparation for birth (RR 2.4, 95% CI 1.0–5.6), self-evaluation checklists (RR 3.8, 95% CI 1.4–9.7) and weekly review and reflection meetings (RR 2.6, 95% 1.0–7.4) helped the health workers to retain their neonatal resuscitation skills. The health workers demonstrated improvement in ventilation of babies within one minute of birth and there was a reduction in intrapartum stillbirth (aOR 0.46, 95% CI 0.32–0.66) and first-day neonatal mortality (aOR 0.51, 95% CI 0.31–0.83).  The study provides information on challenges in reducing stillbirth and neonatal death in low income settings and provides a strategy to improve health workers adherence to neonatal resuscitation to reduce the mortality. The HBB QIC can be implemented in similar clinical settings to improve quality of care and survival in Nepal, but for primary care settings, the QIC need to be evaluated further.

L’analyse visuelle du rythme cardiaque fœtal (RCF) est une excellente méthode de dépistage de l’hypoxie fœtale. Cette analyse visuelle est d’autre part sujette à une variabilité inter- et intra-individuelle importante. L’hypoxie fœtale au cours du travail s’exprime par des anomalies du RCF. La sous-évaluation de la gravité d’un RCF entraine une prise de risque indue pour le fœtus avec une augmentation de sa morbi-mortalité et sa surévaluation entraine un interventionnisme obstétrical inutile avec une augmentation du taux de césariennes. Ce dernier point pose par ailleurs en France un problème de santé publique.L’analyse automatisée du signal RCF permet de diminuer la variabilité inter- et intra-individuelle et d’accéder à d’autres paramètres calculés visant à augmenter la valeur diagnostique. Les critères d’analyse morphologiques du RCF (ligne de base, nombre d’accélérations, nombre et typage des ralentissements, variabilité à long terme (VLT)) ont été décrits ainsi que d’autres tels que les surfaces des ralentissements, les indices de variabilité à court terme (VCT) et les analyses fréquentielles. Il n’en demeure pas moins que la définition de la ligne de base, à partir de laquelle sont repérés les accélérations et les ralentissements reste, dans certains cas, difficile à établir.L’objectif principal de la thèse est d’établir un modèle prédictif de l’acidose fœtale à partir d’une analyse automatisée du RCF. L’objectif secondaire est de déterminer la pertinence des différents paramètres élémentaires classiques (CNGOF 2007) (fréquence de base, variabilité, accélérations, ralentissements) et celle d’autres paramètres inaccessible à l’œil (indices de variabilité à court terme, surfaces des ralentissements, analyse fréquentielle…). Par la suite, nous voulons identifier des critères de décision qui aideront à la prise en charge obstétricale.Nous proposons d’aborder l’analyse automatisée du RCF pendant le travail par l’intermédiaire d’une étude cas-témoins ; les cas étant des tracés RCF de nouveau-nés en acidose néonatale (pH artériel au cordon inférieur ou égal à 7,15) et les témoins, des tracés RCF de nouveau-nés sans acidose (pH artériel au cordon supérieur ou égal à 7,25). Il s’agit d’une étude monocentrique à la maternité de l’hôpital Saint Vincent de Paul, GHICL – Lille, sur notre base de données « Bien Naitre » (archivage numérique des tracés RCF depuis 2011), comptant un un nombre suffisant de cas sur ce seul centre. La maternité Saint Vincent de Paul (GHICL) présente depuis 2011 environ 70 cas par an d’acidose néonatale (pHa ≤ 7,10) (3,41%). Le logiciel R sera utilisé pour l’analyse statistique / Visual analysis of the fetal heart rate FHR is a good method for screening for fetal hypoxia but is not sufficiently specific. The visual morphological analysis of the FHR during labor is subject to inter- and intra-observer variability – particularly when the FHR is abnormal. Underestimating the severity of an FHR leads to undue risk-taking for the fetus with an increase in morbidity and mortality and overvaluation leads to unnecessary obstetric intervention with an increased rate of caesarean section. This last point also induces a French public health problem.FHR automated analysis reduces inter and intra-individual variability and accesses other calculated parameters aimed at increasing the diagnostic value. The FHR morphological analysis parameters (baseline, number of accelerations, number and typing of decelerations, long-term variability (LTV)) were described as well as others such as the decelerations surfaces, short-term variability (STV) and frequency analyzes. Nevertheless, when attempting to analyze the FHR automatically, the main problem is computation of the baseline against which all the other parameters are determined.Automatic analysis provides information on parameters that cannot be derived in a visual analysis and that are likely to improve screening for fetal acidosis during labor.The main objective of the thesis is to establish a predictive model of fetal acidosis from a FHR automated analysis. The secondary objective is to determine the relevance of the classical basic parameters (CNGOF 2007) (baseline, variability, accelerations, decelerations) and that of other parameters inaccessible to the eye (indices of short-term variability, surfaces of decelerations, frequency analysis ...). Later, we want to identify decision criteria that will help in the obstetric care management.We propose to validate FHR automated analysis during labor through a case-control study; cases were FHR recordings of neonatal acidosis (arterial cord pH less than or equal to 7.15) and controls, FHR recordings of neonatal without acidosis (arterial cord pH upper than or equal to 7.25). This is a monocentric study at the maternity hospital of Saint Vincent de Paul Hospital, GHICL - Lille, on our « Well Born » database (digital archiving of RCF plots since 2011), with a sufficient number of cases on this only center. Since 2011, the Saint Vincent de Paul hospital (GHICL) has had about 70 cases per year of neonatal acidosis (pHa less than or equal to 7.10) (3.41%). The R software will be used for statistical analysis.

Αντικείμενο της διατριβής ήταν η ανάπτυξη υπολογιστικών μεθόδων διάγνωσης και εκτίμησης της κατάστασης της υγείας του εμβρύου. Οι προτεινόμενες μεθοδολογίες αναλύουν και εξάγουν πληροφορίες από το σήμα της ΕΚΣ καθώς το συγκεκριμένο σήμα αποτελεί ένα από τα λιγοστά διαθέσιμα εργαλεία για την εκτίμηση της οξυγόνωσης του εμβρύου και της αξιολόγησης της κατάστασης της υγείας του κατά τη διάρκεια του τοκετού. Για την αξιολόγηση των μεθόδων εξετάστηκε η συσχέτιση της Εμβρυϊκής Καρδιακής Συχνότητας (ΕΚΣ) με βραχυπρόθεσμες αξιόπιστες ενδείξεις για την κατάσταση του εμβρύου και πιο συγκεκριμένα χρησιμοποιήθηκε η συσχέτιση της τιμής του pH του αίματος του εμβρύου η οποία αποτελεί μια έμμεση ένδειξη για την ανάπτυξη υποξίας κατά τη διάρκεια του τοκετού. Στα πλαίσια της διατριβής χρησιμοποιήθηκε για πρώτη φορά η μέθοδος της ανάλυσης σε ανεξάρτητες συνιστώσες για την εξαγωγή χαρακτηριστικών από το σήμα της ΕΚΣ. Επίσης προτάθηκαν και χρησιμοποιήθηκαν Κρυφά Μοντέλα Markov σε μια προσπάθεια να «συλληφθεί» η χρονική εξέλιξη του φαινομένου της μεταβολής της κατάστασης του εμβρύου. Επιπλέον προτάθηκαν νέα χαρακτηριστικά εξαγόμενα με τη χρήση του Διακριτού Μετασχηματισμού Κυματιδίου. Με χρήση μιας υβριδική μέθοδος, που βασίζεται στη χρήση εξελικτικής γραμματικής «κατασκευάστηκαν» νέα χαρακτηριστικά παραγόμενα από τα χαρακτηριστικά που είχαν ήδη εξαχθεί με συμβατικές μεθόδους. Επιπρόσθετα στα πλαίσια της διατριβής χρησιμοποιήθηκαν για πρώτη φορά (και η μόνη μέχρι στιγμής) μηχανές διανυσμάτων υποστήριξης για την ταξινόμηση και προτάθηκε και χρησιμοποιήθηκε για πρώτη φορά η μέθοδος βελτιστοποίησης με σμήνος σωματιδίων για τη ρύθμιση των παραμέτρων τους. Τέλος προτάθηκε και χρησιμοποιήθηκε για πρώτη φορά η μέθοδος βελτιστοποίησης με σμήνος σωματιδίων για την εκπαίδευση μιας νέας οικογένειας νευρωνικών δικτύων, των νευρωνικών δικτύων κυματιδίου. Μέσα από τα πειράματα τα οποία διεξήγαμε καταφέραμε να δείξουμε ότι τα δεδομένα της ΕΚΣ διαθέτουν σημαντική πληροφορία η οποία με τη χρήση κατάλληλων προηγμένων μεθόδων επεξεργασίας και ταξινόμησης μπορεί να συσχετιστεί με την τιμή του pH του εμβρύου, κάτι το οποίο θεωρούνταν ουτοπικό στη δεκαετία του 90. / This Dissertation dealt with the development of computational methods for the diagnosis and estimation of fetal condition. The proposed methods analyzed and extracted information from the Fetal Heart Rate (FHR) signal, since this is one of the few available tools for the estimation of fetal oxygenation and the assessment of fetal condition during labor. For the evaluation of the proposed methods the correlation of the FHR signal with short term indices were employed and to be more specific, its correlation with the pH values of fetal blood, which is an indirect sign of the development of fetal hypoxia during labor. In the context of this Dissertation, Independent Component Analysis (ICA) for feature extraction from the FHR signal was used for the first time. Moreover we used Hidden Markov Models in an attempt to “capture” the evolution in time of the fetal condition. Furthermore, new features based on the Discrete Wavelet Transform were proposed and used. Using a new hybrid method based on grammatical evolution new features were constructed based on already extracted features by conventional methods. Moreover, for the first (and only) time, Support Vector Machine (SVM) classifiers were employed in the field of FHR processing and the Particle Swarm Optimization (PSO) method was proposed for tuning their parameters. Finally, a new family of neural networks, the Wavelet Neural Networks (WNN) was proposed and used, trained using the PSO method. By conducting a number of experiments we managed to show that the FHR signal conveys valuable information, which by the use of advanced data processing and classification techniques can be associated with fetal pH, something which was not regarded feasible during the 90’s.

Text in English with abstracts in English, Northern Sotho, Tshivenda and Xitsonga / The purpose of this study was to determine the self-reported competence of newly qualified professional nurses on the critical midwifery skills. The study was conducted in the four public hospitals designated for community service in Gauteng Province. A quantitative descriptive design was used with a structured self-report questionnaire as data collection instrument. Non-probability convenience sampling was used for the study. The sample size was eighty-four newly qualified professional nurses. The Stata 15 software was used for statistical analyses. The researcher used descriptive statistics to describe and synthesize the collected data. The findings revealed that most newly qualified professional nurses had no knowledge in identifying different types of decelerations, management of late and variable decelerations, but had knowledge in most of the skills related to management of third stage of labour. / Maikemišetšo a dinyakišišo tše e be e le go hwetša bokgoni bja go ipega ka nnoši bja baoki bao ba ithutetšego profešene ya booki ka go bokgoni bjo bohlokwa bja pelegišo. Dinyakišišo di dirilwe dipetleleng tše nne tša bohle tšeo di kgethetšwego tirelo ya setšhaba ka Profenseng ya Gauteng. Khwanthitheitif diskriptif disaene ‘Quantitative descriptive design’ e dirišitšwe gammogo le lenaneopotšišo leo le beakantšwego la go ipega ka nnoši ‘structured self-report questionnaire’ bjalo ka sedirišwa sa go kgoboketša bohlatsi. “Non-probability convenience sampling” e dirišitšwe mo go kgetheng banyakišišwa. Bogolo bja sešupo e be e le baoki ba masomeseswai-nne ba baswa bao ba ithutetšego profešene ya booki. “Stata 15 software” e dirišitšwe tshekatshekong ya dipalopalo. Monyakišiši o dirišitše dipalopalo tša tlhalošo ‘descriptive statistics’ go hlaloša le go kopanya ‘data’ yeo e kgobokeditšwego. Ditšweletšo di utollotše gore bontši ba baoki ba baswa bao ba ithutetšego profešene ya booki ga ba na le tsebo ya go hlatha mehuta yeo e fapanego ya diphokotšo, taolo ya diphokotšo tša morago le tša go fetoga, efela ba na le tsebo ka go bokgoni bjo bontši bjoo bo amanago le taolo ya kgato ya boraro ya lešoko. / Ndivho ya ngudo iyi yo vha u wanulusa nḓivho ya vhukoni ha iwe muṋe ya vhaongi vhaswa vha kha ḓi bvaho u phasa vhuongi uri vha na zwikili zwa ndeme zwa vhuongi vhubebisi u swika ngafhi. Ngudo iyi yo itwa kha zwibadela zwiṋa zwa muvhuso zwo ṋewaho u isa tshumelo zwitshavhani kha vunḓu ḽa Gauteng. Kha u kuvhanganya mafhungo muṱoḓisi o shumisa ngona ya u ṱalutshedza ya khwanthithethivi ho ṱanganyiswa na mbudziso dzo dzudzanyiwaho dzi bviselaho khagala kha iwe muṋe (structured self-report questionnaire). Vhunanguludzi ho shumiswaho kha ngudo iyi ho vha “Non-probability convenience”. Tshivhalo tsha vhashelamulenzhe vho nanguludzwaho tsho vha vhaongi vhaswa vha kha ḓibvaho u phasa vha fumalo ina. “The Stata 15 software” ndi tshishumiswa tsho shumiswaho kha u sengulusa mafhungo o kuvhanganywaho. Muṱoḓisisi o shumisa zwisiṱatisitika zwa u ṱalutshedza kha u ṱalutshedza na u dzudzanya mafhungo o kuvhanganyiwaho. Ngudo iyi yo bvisela khagala uri vhunzhi ha vhaongi vhaswa vha kha ḓi bvaho u phasa a vha na nḓivho ya u vhona tshaka dzo fhambanaho dza kurwele kwa mbilu ya ṅwana na u langa u lenga ha u rwa ha mbilu ya ṅwana zwo katela na u sa dzudzanyea fhethu huthihi ha kurwele kwa mbilu ya ṅwana, honeha vha na nḓivho ya zwikili zwi yelanaho na vhulanguli ha tshipiḓa tsha vhuraru tsha u beba. / Xikongomelo xa ndzavisiso lowu i ku kuma vuswikoti lebyi munhu a byi twisisaka hi vaongori lava ha ku thwaselaka tidyondzo ta vuongori eka swikili swa nkoka hi vusungukati. Ndzavisiso lowu wu endliwile eka swibedlhele swa mune swa mani na mani leswi yisaka vukorhokeri evanhwini eka Phurovhinsi ya Gauteng, laha ku tirhisiweke maendlelo ya tinhlayo lama hambanaka na swivutiso ku hlengeleta timhaka. Ku tirhisiwile xiphemu xo karhi xa vanhu ku kuma vuxokoxoko hi mayelano na vona hinkwavo. Xiphemu lexi tirhisiweke i xa nhlayo ya vaongori vo ringana makumenhungu-mune wa vaongori lawa ha ku thwaselaka tidyondzo ta vuongori. Ku tirhisiwile “stata software” ku hlela tinhlayo leti tirhisiweke. Mulavisisi u tirhisile tinhlayo, tinhlayonhlamuselo ku hlamusela no katsakanya mahungu lama a ma hlengeleteke. Leswi kumiweke swi paluxa leswaku vunyingi bya vaongori lava ha ku thwaselaka tidyondzo ta vuongori va hava vutivi byo hambanisa mabelo ya mbilu, ku hlawula ku hlwela no hambana ka mabelo ya mbilu, kambe va na vutivi eka swikili mayelana no lawula xiyimo xa vunharhu xo lumiwa. / Health Studies / M.A. (Nursing)