Related Experiment Video
Updated: Sep 13, 2025
A Modified Sonographic Algorithm for Image Acquisition in Life-Threatening Emergencies in the Critically Ill Newborn
Published on: April 7, 2023
Artificial Intelligence Dystocia Algorithm (AIDA) as a Decision Support System in Transverse Fetal Head Position
Antonio Malvasi1, Lorenzo E Malgieri2, Tommaso Difonzo1
1Unit of Obstetrics and Gynecology, Department of Interdisciplinary Medicine (DIM), University of Bari "Aldo Moro", Policlinic of Bari, Piazza Giulio Cesare 11, 70124 Bari, Italy.
Artificial Intelligence Dystocia Algorithm (AIDA) improves assessment of transverse fetal head position using intrapartum ultrasound. This AI system enhances clinical decisions, reducing cesarean delivery rates by providing objective, geometric parameter-based analysis.
Area of Science:
- Obstetrics and Gynecology
- Medical Imaging
- Artificial Intelligence in Medicine
Background:
- Transverse fetal head position complicates labor, increasing operative delivery and cesarean section rates.
- Traditional digital examination for fetal head position assessment can be unreliable, especially in prolonged labor.
- Intrapartum ultrasound offers better visualization but requires standardized interpretation frameworks.
Purpose of the Study:
- To evaluate the significance of accurate assessment and management of transverse fetal head position.
- To correlate geometric parameters with delivery outcomes using the Artificial Intelligence Dystocia Algorithm (AIDA).
- To analyze AIDA's role as a decision support system for standardizing diagnosis and optimizing management of fetal malposition.
Main Methods:
- Secondary analysis of 66 transverse fetal head position cases from 135 nulliparous women with prolonged second-stage labor.
- Stratification by Midline Angle (MLA) into classic transverse (≥75°), near-transverse (70-74°), and transitional (60-69°) positions.
- Evaluation of four geometric parameters (Angle of Progression, Head-Symphysis Distance, MLA, Asynclitism Degree) using the AIDA classification system and machine learning algorithms (SVM, Random Forest, MLP).
Main Results:
- AIDA categorized labor dystocia into five classes with strong predictive value for delivery outcomes.
- Cesarean delivery risk showed a gradient across transverse positions: near-transverse (100%), classic transverse (93.1%), and transitional (85.7%).
- Random Forest algorithm achieved 95.5% accuracy; concurrent asynclitism significantly increased cesarean delivery rates. All AIDA Class 4 cases required cesarean delivery.
Conclusions:
- AIDA, integrated with intrapartum ultrasound, offers a promising approach for objective assessment and management of transverse fetal head position.
- The AIDA classification system, emphasizing precise MLA measurement, provides superior predictive capability over qualitative assessments.
- This multidimensional AI approach enables personalized, evidence-based management of fetal malpositions, potentially reducing interventions and identifying futile expectant management cases.
