The goal of this clinical trial is to evaluate and compare three digital workflows for the fabrication of definitive implant-supported full-arch prostheses in adult patients requiring fixed implant rehabilitation.

The main questions it aims to answer are:

• Does an automated AI-assisted digital workflow improve the passive fit of definitive full-arch implant-supported prostheses compared with manual and splint-guided alignment workflows?
• Are there differences in marginal, geometric, mechanical, and radiographic passivity among the three digital workflows? Researchers will compare manual CBCT-STL alignment, splint-guided alignment, and automated AI-assisted CBCT-STL alignment to see if the degree of digital workflow automation affects the passive fit of definitive full-arch implant-supported prostheses.

Participants will:

• Be adults (18 years and older) indicated for fixed implant-supported full-arch rehabilitation.
• Receive a definitive, screw-retained, full-arch implant-supported prosthesis fabricated using one of the three assigned digital workflows.
• Undergo standardized clinical and radiographic assessments at the time of definitive prosthesis placement to evaluate prosthesis passive fit.

This randomized controlled clinical trial aims to evaluate and compare the passive fit of definitive full-arch implant-supported prostheses fabricated using three different digital workflows with increasing levels of automation for implant position registration and prosthesis fabrication.

Full-arch implant-supported rehabilitations require a high level of precision to ensure passive fit between the prosthetic framework and the implant-abutment connections. Inadequate passive fit may lead to mechanical complications, biological overload, or long-term prosthetic failure. Digital workflows combining intraoral scanning (IOS) and cone-beam computed tomography (CBCT) have been introduced to improve accuracy; however, differences in data acquisition and dataset alignment strategies may influence the final prosthetic fit.

The study will include adult patients (18 years and older) indicated for fixed implant-supported full-arch rehabilitation in the maxilla or mandible. Eligible participants will be recruited from the Faculty of Dentistry of the Complutense University of Madrid and associated clinical centers. After providing written informed consent, participants will be randomly assigned in a 1:1:1 ratio to one of three digital workflows:

1. Manual CBCT-STL alignment (MedicalFit 1.0): implant positions are obtained by combining intraoral scans and CBCT data, with dataset registration performed manually by the operator based on visual alignment of scan bodies.
2. Splint-guided alignment (MedicalFit 2.0): a calibrated rigid reference splint with metallic cylinders is used to stabilize implant positions and assist manual dataset registration, aiming to reduce operator-dependent variability.
3. Automated AI-assisted CBCT-STL alignment (MedicalFit 3.0 - Pdental): dataset registration and passivation are performed automatically by dedicated software using advanced algorithms for implant detection and alignment, without manual intervention.

All participants will receive a definitive, screw-retained, full-arch implant-supported prosthesis fabricated according to the assigned digital workflow. Prosthetic materials and clinical procedures will follow standard clinical practice. No outcome measures will be assessed during provisional prosthetic phases.

Passive fit will be evaluated exclusively at the time of definitive prosthesis placement using a standardized multi-assessment clinical approach performed at each implant-prosthesis connection.

Marginal passive fit will be assessed clinically by direct inspection using a calibrated periodontal probe. The presence or absence of marginal discrepancies between the prosthesis and the transepithelial abutment will be evaluated based on visual and tactile criteria, including the ability of the probe to penetrate the prosthesis-abutment interface.

Geometric passive fit will be evaluated using the modified Sheffield test. With all prosthetic screws loosened except for one distal screw, the presence of any lifting or separation of the prosthetic framework at the non-tightened connections will be assessed visually and tactually using an explorer probe.

Mechanical passive fit will be assessed through the tactile sensitivity of the passive screw test. Resistance perceived during screw tightening will be evaluated qualitatively to identify the presence of internal stresses or framework flexure during seating of the definitive prosthesis.

Radiographic passive fit will be assessed by measuring marginal gaps at the prosthesis-abutment interface on standardized periapical radiographs. Radiographic gaps will be quantified and classified using predefined ordinal thresholds to identify clinically relevant discrepancies.

Each assessment will be scored using an ordinal scale (0-2) based on predefined clinical criteria. To reflect clinical decision-making and ensure a conservative interpretation of prosthetic fit, a hierarchical aggregation rule will be applied. For each assessment dimension, the prosthesis-level score will be defined as the worst score observed among all implant-prosthesis connections. The global multi-assessment passive fit score (0-2) will then be defined as the worst score observed across all assessment dimensions.

Accordingly, if any implant-prosthesis connection exhibits a clinically relevant misfit in any assessment, the definitive prosthesis will be classified as non-passive. This approach reflects the clinical principle that lack of passive fit at a single connection compromises the overall prosthetic outcome.

The primary objective of the study is to determine whether increased automation in digital workflows improves the passive fit of definitive full-arch implant-supported prostheses. Secondary objectives include comparing marginal, geometric, mechanical, and radiographic passivity outcomes among the three workflows and exploring the relationship between workflow automation and prosthetic adaptation accuracy.

The study follows CONSORT guidelines and aims to provide clinically relevant evidence to support the optimization of digital workflows in full-arch implant-supported prosthetic rehabilitation

Inclusion Criteria:

1. Adults aged 18 years or older.
2. Patients indicated for fixed implant-supported full-arch rehabilitation in the maxilla or mandible.
3. Presence of clinically stable, osteointegrated titanium implants suitable for prosthetic rehabilitation.
4. Absence of clinical or radiographic signs of peri-implant disease.
5. Ability to understand the study procedures and provide written informed consent.
6. Willingness and ability to attend all required clinical visits and evaluations.

Exclusion Criteria:

1. Presence of peri-implant mucositis or peri-implantitis at the time of evaluation.
2. Implant mobility or implant positioning that prevents proper prosthetic rehabilitation.
3. Uncontrolled systemic medical conditions that could interfere with study participation or prosthetic treatment.
4. Cognitive, psychological, or medical conditions that limit the ability to comply with study procedures.
5. Inability to complete the required clinical evaluations or follow the study protocol, as judged by the investigator.

Individual participant data will not be shared due to ethical and privacy considerations. The data collected include detailed clinical and radiographic information that could potentially allow re-identification of participants, even after de-identification. Data will be used solely for the purposes defined in the approved study protocol.