Insights into Predicting Tooth Extraction from Panoramic Dental Images: Artificial Intelligence vs. Dentists.

OBJECTIVES: Tooth extraction is one of the most frequently performed medical procedures. The indication is based on the combination of clinical and radiological examination and individual patient parameters and should be made with great care. However, determining whether a tooth should be extracted is not always a straightforward decision. Moreover, visual and cognitive pitfalls in the analysis of radiographs may lead to incorrect decisions. Artificial intelligence (AI) could be used as a decision support tool to provide a score of tooth extractability. MATERIAL AND METHODS: Using 26,956 single teeth images from 1,184 panoramic radiographs (PANs), we trained a ResNet50 network to classify teeth as either extraction-worthy or preservable. For this purpose, teeth were cropped with different margins from PANs and annotated. The usefulness of the AI-based classification as well that of dentists was evaluated on a test dataset. In addition, the explainability of the best AI model was visualized via a class activation mapping using CAMERAS. RESULTS: The ROC-AUC for the best AI model to discriminate teeth worthy of preservation was 0.901 with 2% margin on dental images. In contrast, the average ROC-AUC for dentists was only 0.797. With a 19.1% tooth extractions prevalence, the AI model's PR-AUC was 0.749, while the dentist evaluation only reached 0.589. CONCLUSION: AI models outperform dentists/specialists in predicting tooth extraction based solely on X-ray images, while the AI performance improves with increasing contextual information. CLINICAL RELEVANCE: AI could help monitor at-risk teeth and reduce errors in indications for extractions.

A 5-year prospective study on single immediate implants in the aesthetic zone.

AIM: There is a paucity of long-term data on soft tissue aesthetics of single immediate implants. The objective of this study was to evaluate the 5-year clinical and aesthetic outcome of this treatment concept. MATERIALS AND METHODS: Twenty-two periodontally healthy patients (12 men, 10 women; mean age 50) with low risk for aesthetic complications (thick gingival biotype, intact buccal bone wall, both neighbouring teeth present) were consecutively treated with a single immediate implant in the aesthetic zone (15-25). Flapless surgery was performed and the gap between the implant and buccal bone wall was systematically filled with bovine bone particles. Implants were immediately non-functionally loaded with a screw-retained provisional crown. Cases demonstrating major alveolar process changes and/or advanced mid-facial recession (>1 mm) at 3 months were additionally treated with a connective tissue graft (CTG). Permanent crowns were installed at 6 months. The clinical and aesthetic results at 5 years were compared to those obtained at 1 year. RESULTS: Seventeen patients attended the 5-year re-assessment, of whom five had been treated with a CTG for early aesthetic complications. There was one early implant failure and one complication after 1 year (porcelain chipping). Mean marginal bone loss was 0.12 mm at 1 year and 0.19 mm at 5 years (p = 0.595) with the moment of implant installation as baseline. Papilla height increased between 1 and 5 years (p ≤ 0.007), whereas mid-facial contour (p = 0.005) and alveolar process deficiency (p = 0.008) deteriorated. Mean mid-facial recession was on average 0.28 mm (SD 0.48) at 1 year and 0.53 mm (SD 0.53) at 5 years (p = 0.072) with the preoperative status as baseline. Three implants demonstrated advanced mid-facial recession (>1 mm) at 5 years. All three were in a central incisor position and none had been treated with a CTG. Thus, 8/17 implants showed aesthetic complications (five early and three late aesthetic complications). Implants in a lateral incisor position showed stable soft tissue levels. The pink aesthetic score was on average 12.15 at 1 year and 11.18 at 5 years (p = 0.030). CONCLUSION: Single immediate implants showed high implant survival and limited marginal bone loss in the long term. However, mid-facial recession, mid-facial contour and alveolar process deficiency deteriorated after 1 year. With an aesthetic complication rate of 8/17 in well-selected patients who had been treated by experienced clinicians, type I placement may not be recommended for daily practice.

Cyto R-CNN and CytoNuke Dataset: Towards reliable whole-cell segmentation in bright-field histological images.

BACKGROUND AND OBJECTIVE: Cell segmentation in bright-field histological slides is a crucial topic in medical image analysis. Having access to accurate segmentation allows researchers to examine the relationship between cellular morphology and clinical observations. Unfortunately, most segmentation methods known today are limited to nuclei and cannot segment the cytoplasm. METHODS: We present a new network architecture Cyto R-CNN that is able to accurately segment whole cells (with both the nucleus and the cytoplasm) in bright-field images. We also present a new dataset CytoNuke, consisting of multiple thousand manual annotations of head and neck squamous cell carcinoma cells. Utilizing this dataset, we compared the performance of Cyto R-CNN to other popular cell segmentation algorithms, including QuPath's built-in algorithm, StarDist, Cellpose and a multi-scale Attention Deeplabv3+. To evaluate segmentation performance, we calculated AP50, AP75 and measured 17 morphological and staining-related features for all detected cells. We compared these measurements to the gold standard of manual segmentation using the Kolmogorov-Smirnov test. RESULTS: Cyto R-CNN achieved an AP50 of 58.65% and an AP75 of 11.56% in whole-cell segmentation, outperforming all other methods (QuPath 19.46/0.91%; StarDist 45.33/2.32%; Cellpose 31.85/5.61%, Deeplabv3+ 3.97/1.01%). Cell features derived from Cyto R-CNN showed the best agreement to the gold standard (D¯=0.15) outperforming QuPath (D¯=0.22), StarDist (D¯=0.25), Cellpose (D¯=0.23) and Deeplabv3+ (D¯=0.33). CONCLUSION: Our newly proposed Cyto R-CNN architecture outperforms current algorithms in whole-cell segmentation while providing more reliable cell measurements than any other model. This could improve digital pathology workflows, potentially leading to improved diagnosis. Moreover, our published dataset can be used to develop further models in the future.