Fetal heart rate pattern interpretation in the second stage of labor using the five-tier classification: Impact of the degree and duration on severe fetal acidosis

Cardiotocography-Based Experimental Comparison of Artificial Intelligence and Human Judgment in Assessing Fetal Asphyxia During Delivery

Cardiotocography (CTG) has long been the standard method for monitoring fetal status during delivery. Despite its widespread use, human error and variability in CTG interpretation contribute to adverse neonatal outcomes, with over 70% of stillbirths, neonatal deaths, and brain injuries potentially avoidable through accurate analysis. Recent advancements in artificial intelligence (AI) offer opportunities to address these challenges by complementing human judgment. This study experimentally compared the diagnostic accuracy of AI and human specialists in predicting fetal asphyxia using CTG data. Machine learning (ML) and deep learning (DL) algorithms were developed and trained on 3,519 CTG datasets. Human specialists independently assessed 50 CTG figures each through web-based questionnaires. A total of 984 CTG figures from singleton pregnancies were evaluated, and outcomes were compared using receiver operating characteristic (ROC) analysis. Human diagnosis achieved the highest area under the curve (AUC) of 0.693 (p = 0.0003), outperforming AI-based methods (ML: AUC = 0.514, p = 0.788; DL: AUC = 0.524, p = 0.662). Although DL-assisted judgment improved sensitivity and identified cases missed by humans, it did not surpass the accuracy of human judgment alone. Combining human and AI predictions yielded a lower AUC (0.693) than human diagnosis alone, but improved specificity (91.92% for humans, 98.03% for humans and DL), highlighting AI's potential to complement human judgment by reducing false-positive rates. Our findings underscore the need for further refinement of AI algorithms and the accumulation of CTG data to enhance diagnostic accuracy. Integrating AI into clinical workflows could reduce human error, optimize resource allocation, and improve neonatal outcomes, particularly in resource-limited settings. These advancements promise a future where AI assists obstetricians in making more objective and accurate decisions during delivery.

Correlation between the 5-tier fetal heart rate pattern classification at delivery and Apgar scores

AIM

In Japan, the fetal heart rate pattern on cardiotocography is usually categorized into five levels for intrapartum management; however, studies on the 5-tier cardiotocography classification are limited. This study aimed to determine the correlation between 5-tier cardiotocography levels at delivery, neonatal Apgar scores and perinatal prognosis using a perinatal registry database.

METHODS

This retrospective study used the 2018 Perinatal Registry Database of the Japan Society of Obstetrics and Gynecology. Apgar scores and perinatal prognosis were compared between patients with Japan Society of Obstetrics and Gynecology intrapartum cardiotocography levels 1-2, and levels 3-5.

RESULTS

In total, 240 987 cases were registered, with 114 201 full-term deliveries comprising 84 703 vaginal and 29 498 cesarean deliveries. For vaginal delivery cases, the crude odds ratios of 1-min Apgar score ≤3 for levels 3 (odds ratio: 1.7, 95% confidence interval: 1.4-2.0, p < 0.0001), levels 4 (odds ratio: 5.9, 95% confidence interval: 5.1-6.9, p < 0.0001), and levels 5 (odds ratio: 49.2, 95% confidence interval: 37.5-64.5, p < 0.0001) were significantly higher than those for levels 1-2 (odds ratio: 1.0). Moreover, the crude odds ratios of 5-min Apgar score ≤ 3 for levels 4 (odds ratio: 4.8, confidence interval: 2.7-8.7, p < 0.0001), and levels 5 (odds ratio: 71.1, confidence interval: 32.6-155.1, p < 0.0001) were significantly higher than those for levels 1-2 (odds ratio: 1.0). Similarly, in cases of cesarean delivery, it showed a similar trend to case of vaginal delivery cases.

CONCLUSION

The 5-tier cardiotocography classification can predict severe neonatal asphyxia.

Accuracy of predicting neonatal distress using a five-level classification of fetal heart rate monitoring

AIM

To assess the accuracy of neonatal distress prediction using the five-level classification of fetal heart rate (FHR) and management protocol of the Japan Society of Obstetrics and Gynecology (JSOG).

METHODS

A case-control study was conducted. Vertex singleton pregnant women who delivered after 37 weeks' gestation from 2013 to 2015 were enrolled. The participants were categorized into two groups; controls were levels 1-3 (n = 1184), whereas cases were levels 4-5 (n = 117) group. Neonatal distress was defined as Apgar score < 8 points at 5 min or umbilical cord artery pH < 7.1.

RESULTS

There were 117 cases (9.0%). The frequency of the neonatal distress was observed in 1.3% controls and 6.8% cases (P < 0.01). Diagnostic accuracy of neonatal distress for cases showed a 6.8% positive-predictive value, 34.8% sensitivity, 91.5% specificity and 98.7% negative-predictive value. Among various obstetrical conditions, high sensitivity (100%) for prediction of neonatal distress was observed in women with chromosome abnormalities, placental abruption, umbilical cord abnormalities and excessive labor pain. Conversely, relatively low specificity (<50%) was observed in cases with oligohydramnios and excessive labor pain.

CONCLUSION

The five-level classification scheme was efficient for neonatal distress prediction. However, depending on the obstetric condition, the FHR findings and neonatal condition might be independent.

[Study of fetal heart rate abnormalities observed on cardiotocography in Lubumbashi: about a cases followed at the Lubumbashi University Clinic and the General Hospital of the Cinquantenaire]

Cardiotocography (CTG) has recently come into use in Lubumbashi but no thorough study has yet been conducted to identify its impact on perinatal morbi-mortality. This study aims to determine the frequency of fetal heart rate abnormalities (FHR)in order to identify the associated factors and to propose a suitable management. We conducted a cross-sectional, descriptive study of 411 women in labour over a period of 19 months (March 2015-December 2016). In patients with pathologic FHR abnormalities, sensitivity and positive predictive value of cardiotocography in the screening test for acute fetal distress were 82.95% and 45.35% respectively. FHR abnormalities were found in two women in labour out of five. Decelerations were the most frequent FHR abnormalities observed (50.8%) with a remarkable predominance of late decelerations (22.1% of all abnormalities). The factors associated with pathological FHR abnormalities were prolonged labor (OR = 14.64, CI = 3.91-54.81), chorioamnionitis (OR = 14.56, CI = 3.83-55.34), chronic maternal anemia (OR = 4.99, CI = 1.48-16.85), primiparity (OR = 2.69, CI = 1.49-4.85), prematurity (OR = 2.90, CI = 1.51-5.54) and prolonged pregnancy (OR = 3.22, CI = 1.38-7.52). Intrauterine growth retardation and arterial hypertension were mainly associated with flat lines and late decelerations (OR = 7.79, CI = 2.50-24.30 and OR=2.74, CI = 1.31-5.72). CTG is a screening tool for the identification of acute fetal distress but with high false-positive rate (55%); it should be associated with other second-line screening tests for acute fetal distress in order to reduce this rate. Factors associated with pathologic FHR abnormalities often cause acute fetal distress thus requiring a rigorous analysis of CTG traces.

Is the neonatal creatine phosphokinase level a reliable marker for fetal hypoxia?

AIM

The creatine phosphokinase (CPK) level is believed to increase in neonatal peripheral blood after tissue damage, including damage from perinatal hypoxia. However, it is not clear whether it is truly a reliable marker for fetal hypoxia. We investigated the chronological changes in neonatal CPK and the reliability of CPK as a marker for fetal hypoxia.

METHODS

Sixty term neonates admitted to the neonatal intensive care unit at Tokyo Women's Medical University Medical Center East from April 2009 to April 2010 were enrolled in this study. We evaluated whether asphyxia and fetal heart rate (FHR) abnormality could predict the neonatal CPK level by using receiver-operator curve analysis. We also compared umbilical cord blood pH levels with neonatal CPK levels. In addition, we investigated factors that influence neonatal CPK in non-asphyxia cases.

RESULTS

The median value of CPK peaked on day 1. There were no significant differences in CPK levels regardless of the presence of asphyxia or FHR abnormality. Non-asphyxiated neonates with older gestational ages and amniotic fluid abnormalities had significantly higher levels of CPK.

CONCLUSION

Our results indicate that the neonatal CPK level is not an appropriate marker for retrospectively predicting either asphyxia or FHR abnormality. There are influencing factors other than asphyxia that increase neonatal CPK. Therefore, one should be careful when making a diagnosis of perinatal hypoxia based solely on increased levels of neonatal CPK after birth.