Automated design prediction for definitive obturator prostheses: A case-based reasoning study

Area of Science:

• Biomedical Engineering
• Dental Prosthetics
• Artificial Intelligence in Healthcare

Background:

• Maxillectomy surgery often requires obturator prostheses to restore function and aesthetics.
• Designing these prostheses can be complex and time-consuming for clinicians.
• Predictive systems can aid in streamlining the design process.

Purpose of the Study:

• To evaluate the effectiveness of a case-based reasoning (CBR) system for predicting obturator prosthesis designs.
• To assess the system's ability to match new maxillectomy cases with historical data.
• To determine the clinical utility of a CBR system in maxillofacial prosthodontics.

Main Methods:

• A database of 209 maxillectomy cases was created using patient images and clinical data.
• A CBR system was developed to identify similar past cases based on defect and prosthesis characteristics.
• Clinicians evaluated the accuracy of prosthesis designs generated by the CBR system for 33 test cases.

Main Results:

• A significant positive correlation (ρ = 0.84, p < 0.0001) was found between system confidence and design accuracy.
• The system achieved a median precision of 0.8 at five cases, indicating effective retrieval of relevant designs.
• Clinician assessments confirmed the system's ability to predict appropriate prosthesis designs.

Conclusions:

• The developed CBR system effectively predicts obturator prosthesis designs for maxillectomy patients.
• The system has the potential to reduce clinician workload and simplify the prosthetic design process.
• Enhanced patient engagement is anticipated through faster insights into prosthetic design options.