This prospective diagnostic study aims to validate the clinical utility of a "Periodontal Panoramic Map" generated by the PerioAI V2.0 system, an artificial intelligence-based platform that integrates intraoral scans and cone-beam CT data, for preoperative diagnosis and surgical planning in patients with moderate to severe periodontitis (Stage II-IV). Current clinical standards-manual probing and two-dimensional radiography-have inherent limitations in accurately visualizing complex three-dimensional bone defect morphology, leading to potential underestimation of disease severity and suboptimal surgical outcomes. Building upon our team's previously published high-precision PerioAI V1.0 system, this study will enroll 80 patients requiring periodontal surgery. Preoperative intraoral scans and cone-beam CT images will be acquired as part of routine care, and the PerioAI V2.0 system will automatically generate a "Periodontal Panoramic Map" with intelligent outputs including probing depth, clinical attachment loss, bone defect morphology classification, furcation involvement grading, and automated measurements of key parameters such as intra-bony defect depth and width. These automated diagnostic results will be compared against the gold standard of full mouth clinical examination and intra-operative direct measurements and observations obtained during periodontal surgery under strict blinded conditions. The primary outcome measures are the accuracy of bone defect morphology classification and the agreement between automated and intra-operative linear measurements assessed by intraclass correlation coefficients and Bland-Altman analysis. Secondary outcomes include accuracy of probing depth, clinical attachment loss, periodontitis staging and grading, furcation involvement grading and treatment planning. This study will provide critical evidence supporting the paradigm shift in periodontal surgery from experience-dependent assessment to data-driven precision medicine, ultimately offering clinicians an intuitive, quantitative, and three-dimensional visualization tool for optimized surgical decision-making.
Aged ≥ 18 years.
Diagnosed with Stage II-IV periodontitis according to the 2018 Classification of Periodontal Diseases.
Presence of at least one tooth requiring periodontal surgery (including open flap debridement or regenerative surgery) due to periodontitis, where the intra-bony defect can be exposed intra-operatively for measurement.
Voluntary participation and provision of written informed consent.
Pregnant or lactating women.
Presence of uncontrolled systemic diseases that significantly affect surgery or tissue healing, such as uncontrolled diabetes mellitus or immunodeficiency.
History of head and neck radiotherapy.
Inability to cooperate with the required study examinations.
Patient diagnosed with Stage II-IV periodontitis and has at least one tooth requiring periodontal surgery
Cohort
Prospective
Perio AI V2.0 System
Diagnostic testThis is a single-arm, prospective diagnostic accuracy study. The intervention is the application of an artificial intelligence-based software (PerioAI V2.0) to routinely acquired preoperative intra-oral scan and cone-beam CT data. The software generates a "Periodontal Panoramic Map" with automated measurements and classifications. All participants then undergo routine full clinical examination and clinically indicated periodontal surgery, which are obtained as the gold standard to validate the accuracy of the Perio AI V2.0 system's preoperative diagnostic outputs. The study does not involve any experimental therapeutic interventions; all surgical procedures are part of standard care.
Accuracy of Bone Defect Morphology Classification by PerioAI System Compared to Intraoperative Findings
The PerioAI 2.0 system automatically classifies bone defect morphology (1-wall, 2-wall, 3-wall intrabony defects, dehiscence, or fenestration) based on preoperative intraoral scan and cone-beam CT data. The classification accuracy is assessed by comparing the PerioAI-generated classification against the gold standard of intraoperative direct visual observation by an experienced surgeon during periodontal surgery. Results are reported as the percentage of correctly classified defects (accuracy rate), with sensitivity and specificity for each defect type.
Agreement Between PerioAI-Automated Probing Depth Measurements and Clinical Probing Depth
Probing depth (PD) is measured in millimeters at six sites per tooth (mesio-buccal, mid-buccal, disto-buccal; mesio-lingual, mid-lingual, disto-lingual). The PerioAI system automatically measures PD from fused intraoral scan and cone-beam CT data. These automated measurements are compared with clinical PD measurements obtained by a trained periodontist using a manual periodontal probe (UNC-15). Agreement is assessed using intraclass correlation coefficients and Bland-Altman analysis.
Agreement Between PerioAI-Automated Clinical Attachment Loss Measurements and Clinical Attachment Loss
Clinical attachment loss (CAL) is measured in millimeters as the distance from the cementoenamel junction to the base of the pocket at six sites per tooth. The PerioAI system automatically measures CAL from fused intraoral scan and cone-beam CT data. These automated measurements are compared with clinical CAL measurements obtained by a trained periodontist using a manual periodontal probe (UNC-15). Agreement is assessed using intraclass correlation coefficients and Bland-Altman analysis.
Agreement Between PerioAI-Automated Periodontitis Staging and Grading and Clinical Staging and Grading
Periodontitis stage (II, III, or IV) and grade (A, B, or C) are determined according to the 2018 Classification of Periodontal Diseases. The PerioAI system automatically assigns stage and grade based on analysis of intraoral scan and cone-beam CT data (including radiographic bone loss, tooth loss, defect complexity, and other parameters). These automated classifications are compared with clinical staging and grading performed by a trained periodontist using full-mouth periodontal charting, radiographic review, and the 2018 classification criteria. Agreement is assessed using weighted kappa statistics.
Accuracy of Furcation Involvement Grading by PerioAI System Compared to Intraoperative Findings
Furcation involvement is graded as Grade I, II, or III according to the 2018 Classification of Periodontal Diseases. The PerioAI system automatically assigns furcation involvement grade based on cone-beam CT data. The accuracy of the PerioAI-generated grading is assessed by comparing it against the gold standard of intraoperative direct exploration using a Nabers probe during periodontal surgery. Results are reported as the percentage of correctly graded furcations (accuracy rate), with sensitivity and specificity for each grade.
Click on the lead sponsor to view all of their trials.
This trial is not recruiting at the moment. You can still explore other options:
