Postoperative pain is a common and significant problem following open heart surgery. Fascial plane blocks (FPBs) such as Serratus Anterior Plane Block (SAPB), Pectoral Nerve Blocks (PECS), and Pecto-Intercostal Fascial Block (PIFB) are increasingly used as part of multimodal analgesia in cardiac surgery. However, objective assessment of block quality and its relationship with clinical outcomes remains limited in the literature.

This prospective observational study aims to evaluate the anatomical success of ultrasound-guided fascial plane blocks applied in elective open heart surgery (median sternotomy) using two simultaneous methods: a trained artificial intelligence (AI) model and a blinded expert anesthesiologist. Block images will be recorded in DICOM format and scored on a 3-point scale (1: incorrect anatomical placement, 2: patchy spread, 3: ideal anatomical placement). The relationship between anatomical block success scores and postoperative pain (NRS at 0, 6, 12, 24, and 48 hours), total analgesic consumption, and clinical outcomes will be investigated. Agreement between AI and blinded anesthesiologist assessments will also be analyzed.

Patients scheduled for elective open heart surgery via median sternotomy will be enrolled after obtaining written informed consent, including separate consent for AI-based image analysis. All patients will undergo routine anesthetic management including BIS monitoring, intra-arterial cannulation, standard anesthesia induction, orotracheal intubation, and central venous catheterization.

Fascial Plane Block Application:

Following general anesthesia induction, fascial plane blocks routinely performed in our clinic will be applied by a blinded anesthesiologist using a high-frequency linear ultrasound probe (10-15 MHz, G-brand). The blocks applied include SAPB + PIFB combination or bilateral parasternal (PIFB) block, depending on the surgical decision. All blocks will be performed using 0.25% bupivacaine (total bilaterally 30 mL) under sterile conditions with an in-plane technique.

Image Recording Protocol:

Block images will be recorded in a standardized manner including: pre-block anatomical scanning, video recording during needle placement, local anesthetic spread images, and post-injection final images. All images will be recorded in DICOM format (minimum 1920x1080 resolution) with standardized depth, gain, and focus settings. Patient identifiers will be anonymized.

AI Model Development:

A Convolutional Neural Network (CNN)-based model (U-Net or ResNet architecture) will be trained using expert-labeled ultrasound images from NYSORA references. The dataset will be divided into 70% training, 20% validation, and 10% test sets using data augmentation and cross-validation techniques. The model will evaluate anatomical landmark recognition accuracy, local anesthetic spread segmentation (Dice coefficient), and similarity to ideal block (0-100 score).

Block Quality Scoring:

Recorded ultrasound images will be independently scored by both the trained AI model and a blinded expert anesthesiologist using a 3-point scale:

1. Anatomically incorrect block
2. Anatomically patchy block
3. Anatomically ideal block

Postoperative Follow-up:

Patients will be followed postoperatively with NRS pain scores (at rest and with coughing) at 0, 6, 12, 24. Additional parameters recorded include total analgesic consumption, side effects (nausea, vomiting, pruritus), complications, mobilization time, ICU stay, and hospital length of stay.

Statistical Analysis:

Correlation between AI scores and postoperative pain scores will be assessed using Pearson or Spearman correlation. Factors affecting pain scores will be evaluated by multiple linear regression. Differences between block types will be analyzed using ANOVA or Kruskal-Wallis test. Agreement between AI and blinded anesthesiologist assessments will be analyzed using Cohen's kappa coefficient.

Sample Size:

Based on pilot study data, 120 patients are planned with 95% power.

Inclusion Criteria:

• Age between 18-80 years
• Patients scheduled for elective cardiac surgery (CABG, valve surgery, combined procedures)
• Patients undergoing median sternotomy
• ASA physical status II-III
• Patients capable of providing informed consent, including separate consent for AI-based image analysis

Exclusion Criteria:

• Emergency surgery
• Redo sternotomy
• Allergy to local anesthetics
• Coagulopathy (INR >1.5, platelet count <80,000/mm³)
• Infection at the block application site
• Chronic opioid use (>3 months)
• Cognitive dysfunction
• Pregnancy
• ASA physical status IV and above
• Minimally invasive surgery performed via thoracotomy

Patients scheduled for elective open heart surgery via median sternotomy at the University of Health Sciences Bursa Yüksek Ihtisas Training and Research Hospital, Department of Anesthesiology and Reanimation. All eligible patients who meet the inclusion criteria and provide written informed consent, including separate consent for AI-based ultrasound image analysis, will be enrolled consecutively.

Individual participant data will not be shared due to patient privacy and confidentiality concerns, as the study involves anonymized ultrasound images and clinical data collected at a single center. Data may be made available upon reasonable request to the principal investigator.