Use of Convolutional Neural Networks for Detection of Pathologies in Dental X-Ray Images in Clinical Decision Support Systems
Bibliographic record
Abstract
Relevance: The relevance of the study is determined by the need for automated, scalable solutions for processing large volumes of dental radiological images, which provide precise segmentation, detection, and classification of pathologies in the integrated Clinical Decision Support System (CDSS) modules. Aim: The aim of the study is to develop, optimize, and verify a HITL-CDSS framework for dental radiology with multi-level integration of Convolutional Neural Network (CNN) models, ensuring architectural consistency, metric validity, and expert adaptability. Methods: Research methods: critical architectural and functional analysis of CNN models, metric and indicator modelling of efficiency, synthesis and Unified Modelling Language-based (UML)modelling of the CDSS framework, UML optimization with Human-in-the-loop (HITL) integration, metric and indicator verification of HITL-CDSS. Results: Architectural and functional, metric and indicator, as well as UML modelling of CNN architectures was carried out for the purpose of integration into the dental radiology CDSS. The resultant HITL-optimized framework based on DenseNet/EfficientNet, HRNet, YOLOv8 provided AUC = 0.96–0.98, F1@t = 0.91–0.94, DSC = 0.89–0.92, mAP = 0.72–0.77 at ECE = 0.02–0.04. Integration of HITL mechanisms increased Explainable Artificial Intelligence (XAI) interpretability, resistance to domain shifting, and clinical validity, indicating the appropriateness of multi-modular construction of CDSS with the inclusion of expert feedback. Conclusion: The academic novelty of the study is the development of a HITL-CDSS framework with multi-level CNN integration, which provides metrically verified interpretability, domain-stable generalizability, and clinical relevance in dental radiology tasks.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".