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.

Deep-learning based classification of a tumor marker for prognosis on Hodgkin's disease.

PURPOSE: Hodgkin's disease is a common malignant disorder in adolescent patients. Although most patients are cured, approximately 10%-15% of patients experience a relapse or have resistant disease. Furthermore, there are no definitive molecular predictors for early identification of patients at high risk of treatment failure to first line therapy. The aim of this study was to evaluate the deep learning-based classifier model of medical image classification to predict clinical outcome that may help in appropriate therapeutic decisions. METHODS: Eighty-three FFPE biopsy specimens from patients with Hodgkin's disease were stratified according to the patient's qPET scores, stained with picrosirius red dye and digitalized by whole slide image scanning. The resulting whole slide images were cut into tiles and annotated by two classes based on the collagen fibers' degree of coloring with picrosirius red. The neural network (YOLOv4) was then trained with the annotated data. Training was performed with 30 cases. Prognostic power of the weakly stained picrosirius red fibers was evaluated with 53 cases. The same neural network was trained with MMP9 stained tissue slides from the same cases and the quantification results were compared with the variant from the picrosirius red cases. RESULTS: There was a weak monotonically increasing relationship by parametric ANOVA between the qPET groups and the percentages of weakly stained fibers (p = .0185). The qPET-positive cases showed an average of 18% of weakly stained fibers, and the qPET-negative cases 10%-14%. Detection performance showed an AUC of 0.79. CONCLUSIONS: Picrosirius red shows distinct associations as a prognostic metric candidate of disease progression in Hodgkin's disease cases using whole slide images but not sufficiently as a prognostic device.

Advantages of a Training Course for Surgical Planning in Virtual Reality for Oral and Maxillofacial Surgery: Crossover Study.

BACKGROUND: As an integral part of computer-assisted surgery, virtual surgical planning (VSP) leads to significantly better surgery results, such as for oral and maxillofacial reconstruction with microvascular grafts of the fibula or iliac crest. It is performed on a 2D computer desktop screen (DS) based on preoperative medical imaging. However, in this environment, VSP is associated with shortcomings, such as a time-consuming planning process and the requirement of a learning process. Therefore, a virtual reality (VR)-based VSP application has great potential to reduce or even overcome these shortcomings due to the benefits of visuospatial vision, bimanual interaction, and full immersion. However, the efficacy of such a VR environment has not yet been investigated. OBJECTIVE: This study aimed to demonstrate the possible advantages of a VR environment through a substep of VSP, specifically the segmentation of the fibula (calf bone) and os coxae (hip bone), by conducting a training course in both DS and VR environments and comparing the results. METHODS: During the training course, 6 novices were taught how to use a software application in a DS environment (3D Slicer) and in a VR environment (Elucis) for the segmentation of the fibula and os coxae, and they were asked to carry out the maneuvers as accurately and quickly as possible. Overall, 13 fibula and 13 os coxae were segmented for each participant in both methods (VR and DS), resulting in 156 different models (78 fibula and 78 os coxae) per method (VR and DS) and 312 models in total. The individual learning processes in both environments were compared using objective criteria (time and segmentation performance) and self-reported questionnaires. The models resulting from the segmentation were compared mathematically (Hausdorff distance and Dice coefficient) and evaluated by 2 experienced radiologists in a blinded manner. RESULTS: A much faster learning curve was observed for the VR environment than the DS environment (ß=.86 vs ß=.25). This nearly doubled the segmentation speed (cm3/min) by the end of training, leading to a shorter time (P<.001) to reach a qualitative result. However, there was no qualitative difference between the models for VR and DS (P=.99). The VR environment was perceived by participants as more intuitive and less exhausting, and was favored over the DS environment. CONCLUSIONS: The more rapid learning process and the ability to work faster in the VR environment could save time and reduce the VSP workload, providing certain advantages over the DS environment.

Tolerable degree of muscle sacrifice when harvesting a vastus lateralis or myocutaneous anterolateral thigh flap.

The myocutaneous anterolateral thigh (ALT) and vastus lateralis (VL) flaps include a large muscle mass and a sufficient vascular pedicle, and they have been used for decades to reconstruct traumatic and acquired defects of the head and neck and extremities. In spite of these benefits, musculoskeletal dysfunction was reported in nearly 1 out of 20 patients at follow-up. It is unclear whether the recently proposed muscle-sparing flap-raising approach could preserve VL muscle function and whether patients at increased risk could benefit from such an approach. Therefore, we performed a predictive dynamic gait simulation based on a biological motion model with gradual weakening of the VL during a self-selected and fast walking speed to determine the compensable degree of VL muscle reduction. Muscle force, joint angle, and joint moment were measured. Our study showed that VL muscle reduction could be compensated up to a certain degree, which could explain the observed incidence of musculoskeletal dysfunction. In elderly or fragile patients, the VL muscle should not be reduced by 50% or more, which could be achieved by muscle-sparing flap-raising of the superficial partition only. In young or athletic patients, a VL muscle reduction of 10%, which corresponds to a muscle cuff, has no relevant effect. Yet, a reduction of more than 30% leads to relevant weakening of the quadriceps. Therefore, in this patient population with the need for a large portion of muscle, alternative flaps should be considered. This study can serve as the first basis for further investigations of human locomotion after flap-raising.

Marginal quality and wear of extended posterior resin composite restorations: Eight-year results in vivo.

AIM: To evaluate wear and marginal quality of resin composite restorations over eight years of clinical service in vivo. METHODS: 30 patients received 68 resin composite restorations (36 Grandio, 32 Tetric Ceram) in the course of a prospective clinical trial. 3-D evaluation of 36 selected teeth involving 144 epoxy replicas was carried out using a special 3-D scanning device with an accuracy of U1=2.5+L/350 µm; U3=3+L/300 µm. In vivo replicas were sputter-coated with gold and examined under a SEM at 200× magnification. Marginal integrity between resin composite and enamel was expressed as a percentage of the entire judgeable margin length. RESULTS: During the clinical 8-year observation period, wear significantly increased in the restored areas as well as in OCAs. RBCs under investigation showed no significant differences regarding wear (p>0.05). Localization of the restorations (premolar vs. molar or upper vs. lower) did not show a significant influence on wear rates (p>0.05). Clinically, by SEM, and by 3-D scanning distinct changes of worn contours on enamel and RBC were visible. Quantitative margin analysis revealed a change of perfect margins (58% at baseline vs. 14% at 8 years), positive step formations (15% at baseline vs. 10% at 8 years), and negative step formations (20% at baseline vs. 71% at 8 years) over time (p<0.001). Regarding the portion of gap-free margins (baseline vs. 2 years) and total margin length (baseline to 8 years), Grandio showed lower values (Mann-Whitney U-test; p<0.05). The portion of negative step formations was lower for Tetric Ceram at baseline (Mann-Whitney-U test; p<0.05). CONCLUSIONS: After eight years of clinical service, neither wear nor marginal quality was a critical factor for estimation and survival of extended posterior resin composite restorations.