2023

Jump to: 2022, 2021, 2020

11 pages, 3321 KiB

Open AccessArticle

Evaluation of Intraoral Full-Arch Scan versus Conventional Preliminary Impression

by Kinga Mária Jánosi, Diana Cerghizan, Krisztina Ildikó Mártha, Éva Elekes, Brigitta Szakács, Zoltán Elekes, Alpár Kovács, Andrea Szász, Izabella Mureșan and Liana Georgiana Hănțoiu

J. Clin. Med. 2023, 12(17), 5508; https://doi.org/10.3390/jcm12175508 - 24 Aug 2023

Cited by 1 | Viewed by 1087

Abstract

An accurate impression is vital during prosthodontic rehabilitation. Digital scanning has become an alternative to conventional impressions. This study compares conventional preliminary impression techniques with digital scanning, evaluating the efficiency, treatment comfort, and trueness. Impressions of 28 patients were taken using conventional and [...] Read more.

An accurate impression is vital during prosthodontic rehabilitation. Digital scanning has become an alternative to conventional impressions. This study compares conventional preliminary impression techniques with digital scanning, evaluating the efficiency, treatment comfort, and trueness. Impressions of 28 patients were taken using conventional and digital techniques. The efficiency of both impression techniques was evaluated by measuring the mean working time. A visual analog scale questionnaire (1–10) was used to appreciate the participants’ perceptions of discomfort. Morphometric measurements, which were carried out to determine the differences between the casts, were made on the buccolingual cross sections of teeth 11 and 31 and the distolingual and mesiobuccal cusp tips of each first molar. The total treatment time was 75.5 min for conventional and 12 min for digital impressions. The patients scored a mean discomfort assessment of 6.66 for conventional and 9.03 for digital scanning. No significant differences existed between the examined areas (p < 0.05, Wilcoxon and Mann–Whitney tests) of the digital casts obtained by both techniques. The intraoral scan can be considered as an alternative to conventional preliminary impressions for performing study model analysis during orthodontic treatment planning. The digital impression is more comfortable and accepted by the patients than the conventional impression and has a shorter working time. Full article

► Show Figures

Graphical abstract

2022

Jump to: 2023, 2021, 2020

9 pages, 894 KiB

Open AccessSystematic Review

Accuracy of Dynamic Navigation for Non-Surgical Endodontic Treatment: A Systematic Review

by Egle Marija Jonaityte, Goda Bilvinaite, Saulius Drukteinis and Andres Torres

J. Clin. Med. 2022, 11(12), 3441; https://doi.org/10.3390/jcm11123441 - 15 Jun 2022

Cited by 6 | Viewed by 2329

Abstract

In recent years, the application of Guided Endodontics has gained interest for non-surgical endodontic treatment and retreatment. The newest research focuses on the accuracy of Dynamic Navigation (DN). This article systematically reviewed existing data on the accuracy of non-surgical endodontic treatment procedures that [...] Read more.

In recent years, the application of Guided Endodontics has gained interest for non-surgical endodontic treatment and retreatment. The newest research focuses on the accuracy of Dynamic Navigation (DN). This article systematically reviewed existing data on the accuracy of non-surgical endodontic treatment procedures that were completed using DN. Following the PRISMA criteria, an electronic database search was conducted in PubMed, Web of Science, Scopus, and Cochrane Library. Studies comparing the accuracy of non-surgical endodontic treatment using DN and the conventional freehand technique were eligible. The literature search resulted in 176 preliminary records. After the selection process six studies were included. The risk of bias was evaluated using the modified Cochrane Collaboration Risk of Bias 2.0 tool. Five studies examined the aid of DN for planning and executing endodontic access cavities, and one for fiber post removal. In two studies, endodontic access cavities were performed in teeth with pulp canal obliteration. The main outcomes that were measured in the included studies were preparation time, global coronal entry point and apical endpoint deviations, angular deviation, tooth substance loss, qualitative precision, number of unsuccessful attempts or procedural mishaps. The risk of bias was rated from low to raising some concerns. Overall, DN showed increased accuracy compared to the freehanded technique and could be especially helpful in treating highly difficult endodontic cases. Clinical studies are needed to confirm the published in vitro data. Full article

► Show Figures

Figure 1

13 pages, 2916 KiB

Open AccessArticle

New Methods for the Acoustic-Signal Segmentation of the Temporomandibular Joint

by Marcin Kajor, Dariusz Kucharski, Justyna Grochala and Jolanta E. Loster

J. Clin. Med. 2022, 11(10), 2706; https://doi.org/10.3390/jcm11102706 - 11 May 2022

Cited by 2 | Viewed by 1491

Abstract

(1) Background: The stethoscope is one of the main accessory tools in the diagnosis of temporomandibular joint disorders (TMD). However, the clinical auscultation of the masticatory system still lacks computer-aided support, which would decrease the time needed for each diagnosis. This can be [...] Read more.

(1) Background: The stethoscope is one of the main accessory tools in the diagnosis of temporomandibular joint disorders (TMD). However, the clinical auscultation of the masticatory system still lacks computer-aided support, which would decrease the time needed for each diagnosis. This can be achieved with digital signal processing and classification algorithms. The segmentation of acoustic signals is usually the first step in many sound processing methodologies. We postulate that it is possible to implement the automatic segmentation of the acoustic signals of the temporomandibular joint (TMJ), which can contribute to the development of advanced TMD classification algorithms. (2) Methods: In this paper, we compare two different methods for the segmentation of TMJ sounds which are used in diagnosis of the masticatory system. The first method is based solely on digital signal processing (DSP) and includes filtering and envelope calculation. The second method takes advantage of a deep learning approach established on a U-Net neural network, combined with long short-term memory (LSTM) architecture. (3) Results: Both developed methods were validated against our own TMJ sound database created from the signals recorded with an electronic stethoscope during a clinical diagnostic trail of TMJ. The Dice score of the DSP method was 0.86 and the sensitivity was 0.91; for the deep learning approach, Dice score was 0.85 and there was a sensitivity of 0.98. (4) Conclusions: The presented results indicate that with the use of signal processing and deep learning, it is possible to automatically segment the TMJ sounds into sections of diagnostic value. Such methods can provide representative data for the development of TMD classification algorithms. Full article

► Show Figures

Figure 1

13 pages, 4404 KiB

Open AccessArticle

Comparison of Two 3D-Printed Indirect Bonding (IDB) Tray Design Versions and Their Influence on the Transfer Accuracy

by Julius von Glasenapp, Eva Hofmann, Julia Süpple, Paul-Georg Jost-Brinkmann and Petra Julia Koch

J. Clin. Med. 2022, 11(5), 1295; https://doi.org/10.3390/jcm11051295 - 26 Feb 2022

Cited by 6 | Viewed by 3479

Abstract

Objective: This study aims to investigate the transfer accuracy of two different design versions for 3D-printed indirect bonding (IDB) trays. Materials and Methods: Digital plaster models of 27 patients virtually received vestibular attachments on every tooth using OnyxCeph³™ (Image Instruments, Chemnitz, Germany). Based [...] Read more.

Objective: This study aims to investigate the transfer accuracy of two different design versions for 3D-printed indirect bonding (IDB) trays. Materials and Methods: Digital plaster models of 27 patients virtually received vestibular attachments on every tooth using OnyxCeph³™ (Image Instruments, Chemnitz, Germany). Based on these simulated bracket and tube positions, two versions of transfer trays were designed for each dental arch and patient, which differed in the mechanism of bracket retention: Variant one (V1) had arm-like structures protruding from the tray base and reaching into the horizontal and vertical bracket slots, and variant two (V2) had a pocket-shaped design enclosing the brackets from three sides. Both tray designs were 3D-printed with the same digital light processing (DLP) printer using a flexible resin-based material (IMPRIMO^® LC IBT/Asiga MAX™, SCHEU-DENTAL, Iserlohn, Germany). Brackets and tubes (discovery^® smart/pearl, Ortho-Cast M-Series, Dentaurum, Ispringen, Germany) were inserted into the respective retention mechanism of the trays and IDB was performed on corresponding plaster models. An intraoral scan (TRIOS^® 3W, 3Shape, Copenhagen, Denmark) was performed to capture the actual attachment positions and compared to the virtually planned positions with Geomagic^© Control (3D Systems Inc., Rock Hill, SC, USA) using a scripted calculation tool, which superimposed the respective tooth surfaces. The resulting attachment deviations were determined in three linear (mesiodistal, vertical and orovestibular) and three angular (torque, rotation and tip) directions and analyzed with a descriptive statistical analysis. A comparison between the two IDB tray designs was conducted using a mixed model analysis (IBM, SPSS^® Statistics 27, Armonk, NY, USA). Results: Both design versions of the 3D-printed IDB trays did not differ significantly in their transfer accuracy (p > 0.05). In total, 98% (V1) and 98.5% (V2) of the linear deviations were within the clinically acceptable range of ±0.2 mm. For the angular deviations, 84.9% (V1) and 86.8% (V2) were within the range of ±1°. With V1, most deviations occurred in the mesiodistal direction (3.3%) and in rotation (18%). With V2, most deviations occurred in the vertical direction (3.8%) and in palatinal and lingual crown torque (16.3%). Conclusions: The transfer accuracies of the investigated design versions for 3D-printed IDB trays show good and comparable results albeit their different retention mechanisms for the attachments and are, therefore, both suitable for clinical practice. Full article

► Show Figures

Figure 1

10 pages, 17967 KiB

Open AccessArticle

Consistency and Reliability Analyses of a Comprehensive Index for the Evaluation of Teeth Alignment Performance

by Andrea Mapelli, Marco Serafin, Carolina Dolci, Daniele Gibelli, Alberto Caprioglio, Chiarella Sforza and Gianluca Martino Tartaglia

J. Clin. Med. 2022, 11(4), 1016; https://doi.org/10.3390/jcm11041016 - 16 Feb 2022

Cited by 1 | Viewed by 1701

Abstract

(1) Introduction: The purpose of this work was to describe a method and propose a novel accuracy index to assess orthodontic alignment performance. (2) Methods: Fifteen patients who underwent orthodontic treatment using directly printed clear aligners were recruited. The study sample included 12 [...] Read more.

(1) Introduction: The purpose of this work was to describe a method and propose a novel accuracy index to assess orthodontic alignment performance. (2) Methods: Fifteen patients who underwent orthodontic treatment using directly printed clear aligners were recruited. The study sample included 12 maxillary and 10 mandibular arches, whose pre-treatment, predicted and post-treatment digital models were superimposed on the untreated posterior teeth by means of a best-fit surface-based registration, which was also used to transfer three anatomical landmarks, digitally labeled on the crown of each anterior moving tooth, from the pre-treatment to the predicted and post-treatment models. The Teeth Alignment Performance (TAP) index, quantifying how close the final landmarks were to their expected final position, was proposed as an accuracy index of both individual tooth and group of teeth movement, and its inter-examiner repeatability was tested. (3) Results: No systematic inter-rater discrepancy associated with TAP was observed (p > 0.05), not even when a slight systematic inter-rater difference in landmark labelling was detected (for the upper central incisors, p < 0.001). In addition, all Intra-class Correlation Coefficient (ICC) values showed excellent inter-rater agreement (>0.95), and the small Random Error of Measurement (REM), ranging from 1% for the arch TAP to 3% for the lower canine TAP, indicated that this accuracy index is highly repeatable. (4) Conclusions: The TAP index was proven to be comprehensive, consistent and reliable in assessing the performance of teeth alignment according to a digital plan. The proposed method is also suitable to be implemented in the clinical digital workflow. Full article

► Show Figures

Figure 1

12 pages, 7105 KiB

Open AccessArticle

Evaluating the Precision of Automatic Segmentation of Teeth, Gingiva and Facial Landmarks for 2D Digital Smile Design Using Real-Time Instance Segmentation Network

by Seulgi Lee and Jong-Eun Kim

J. Clin. Med. 2022, 11(3), 852; https://doi.org/10.3390/jcm11030852 - 06 Feb 2022

Cited by 7 | Viewed by 3321

Abstract

Digital smile design (DSD) technology, which takes pictures of patients’ faces together with anterior dentition and uses them for prosthesis design, has been recently introduced. However, the limitation of DSD is that it evaluates a patient with only one photograph taken in a [...] Read more.

Digital smile design (DSD) technology, which takes pictures of patients’ faces together with anterior dentition and uses them for prosthesis design, has been recently introduced. However, the limitation of DSD is that it evaluates a patient with only one photograph taken in a still state, and the patient’s profile cannot be observed from various viewpoints. Therefore, this study aims to segment the patient’s anterior teeth, gingiva and facial landmarks using YOLACT++. We trained YOLACT++ on the annotated data of the teeth, lips and gingiva from the Flickr-Faces-HQ (FFHQ) data. We evaluated that the model trained by 2D candid facial images for the detection and segmentation of smile characteristics. The results show the possibility of an automated smile characteristic identification system for the automatic and accurate quantitative assessment of a patient’s smile. Full article

► Show Figures

Figure 1

2021

Jump to: 2023, 2022, 2020

12 pages, 2592 KiB

Open AccessArticle

Estimating Cervical Vertebral Maturation with a Lateral Cephalogram Using the Convolutional Neural Network

by Eun-Gyeong Kim, Il-Seok Oh, Jeong-Eun So, Junhyeok Kang, Van Nhat Thang Le, Min-Kyung Tak and Dae-Woo Lee

J. Clin. Med. 2021, 10(22), 5400; https://doi.org/10.3390/jcm10225400 - 19 Nov 2021

Cited by 18 | Viewed by 3898

Abstract

Recently, the estimation of bone maturation using deep learning has been actively conducted. However, many studies have considered hand–wrist radiographs, while a few studies have focused on estimating cervical vertebral maturation (CVM) using lateral cephalograms. This study proposes the use of deep learning [...] Read more.

Recently, the estimation of bone maturation using deep learning has been actively conducted. However, many studies have considered hand–wrist radiographs, while a few studies have focused on estimating cervical vertebral maturation (CVM) using lateral cephalograms. This study proposes the use of deep learning models for estimating CVM from lateral cephalograms. As the second, third, and fourth cervical vertebral regions (denoted as C2, C3, and C4, respectively) are considerably smaller than the whole image, we propose a stepwise segmentation-based model that focuses on the C2–C4 regions. We propose three convolutional neural network-based classification models: a one-step model with only CVM classification, a two-step model with region of interest (ROI) detection and CVM classification, and a three-step model with ROI detection, cervical segmentation, and CVM classification. Our dataset contains 600 lateral cephalogram images, comprising six classes with 100 images each. The three-step segmentation-based model produced the best accuracy (62.5%) compared to the models that were not segmentation-based. Full article

► Show Figures

Figure 1

17 pages, 4350 KiB

Open AccessArticle

A Complete Digital Workflow for Planning, Simulation, and Evaluation in Orthognathic Surgery

by Sang-Jeong Lee, Ji-Yong Yoo, Sang-Yoon Woo, Hoon Joo Yang, Jo-eun Kim, Kyung-Hoe Huh, Sam-Sun Lee, Min-Suk Heo, Soon Jung Hwang and Won-Jin Yi

J. Clin. Med. 2021, 10(17), 4000; https://doi.org/10.3390/jcm10174000 - 03 Sep 2021

Cited by 6 | Viewed by 3077

Abstract

The purpose of this study was to develop a complete digital workflow for planning, simulation, and evaluation for orthognathic surgery based on 3D digital natural head position reproduction, a cloud-based collaboration platform, and 3D landmark-based evaluation. We included 24 patients who underwent bimaxillary [...] Read more.

The purpose of this study was to develop a complete digital workflow for planning, simulation, and evaluation for orthognathic surgery based on 3D digital natural head position reproduction, a cloud-based collaboration platform, and 3D landmark-based evaluation. We included 24 patients who underwent bimaxillary orthognathic surgery. Surgeons and engineers could share the massive image data immediately and conveniently and collaborate closely in surgical planning and simulation using a cloud-based platform. The digital surgical splint could be optimized for a specific patient before or after the physical fabrication of 3D printing splints through close collaboration. The surgical accuracy was evaluated comprehensively via the translational (linear) and rotational (angular) discrepancies between identical 3D landmarks on the simulation and postoperative computed tomography (CT) models. The means of the absolute linear discrepancy at eight tooth landmarks were 0.61 ± 0.55, 0.86 ± 0.68, and 1.00 ± 0.79 mm in left–right, advance–setback, and impaction–elongation directions, respectively, and 1.67 mm in the root mean square direction. The linear discrepancy in the left–right direction was significantly different from the other two directions as shown by analysis of variance (ANOVA, p < 0.05). The means of the absolute angular discrepancies were 1.43 ± 1.06°, 0.50 ± 0.31°, and 0.58 ± 0.41° in the pitch, roll, and yaw orientations, respectively. The angular discrepancy in the pitch orientation was significantly different from the other two orientations (ANOVA, p < 0.05). The complete digital workflow that we developed for orthognathic patients provides efficient and streamlined procedures for orthognathic surgery and shows high surgical accuracy with efficient image data sharing and close collaboration. Full article

► Show Figures

Figure 1

11 pages, 8117 KiB

Open AccessArticle

Comparison of Deep Learning Models for Cervical Vertebral Maturation Stage Classification on Lateral Cephalometric Radiographs

by Hyejun Seo, JaeJoon Hwang, Taesung Jeong and Jonghyun Shin

J. Clin. Med. 2021, 10(16), 3591; https://doi.org/10.3390/jcm10163591 - 15 Aug 2021

Cited by 28 | Viewed by 3806

Abstract

The purpose of this study is to evaluate and compare the performance of six state-of-the-art convolutional neural network (CNN)-based deep learning models for cervical vertebral maturation (CVM) on lateral cephalometric radiographs, and implement visualization of CVM classification for each model using gradient-weighted class [...] Read more.

The purpose of this study is to evaluate and compare the performance of six state-of-the-art convolutional neural network (CNN)-based deep learning models for cervical vertebral maturation (CVM) on lateral cephalometric radiographs, and implement visualization of CVM classification for each model using gradient-weighted class activation map (Grad-CAM) technology. A total of 600 lateral cephalometric radiographs obtained from patients aged 6–19 years between 2013 and 2020 in Pusan National University Dental Hospital were used in this study. ResNet-18, MobileNet-v2, ResNet-50, ResNet-101, Inception-v3, and Inception-ResNet-v2 were tested to determine the optimal pre-trained network architecture. Multi-class classification metrics, accuracy, recall, precision, F1-score, and area under the curve (AUC) values from the receiver operating characteristic (ROC) curve were used to evaluate the performance of the models. All deep learning models demonstrated more than 90% accuracy, with Inception-ResNet-v2 performing the best, relatively. In addition, visualizing each deep learning model using Grad-CAM led to a primary focus on the cervical vertebrae and surrounding structures. The use of these deep learning models in clinical practice will facilitate dental practitioners in making accurate diagnoses and treatment plans. Full article

► Show Figures

Figure 1

10 pages, 747 KiB

Open AccessArticle

Factors Influencing Patient Compliance during Clear Aligner Therapy: A Retrospective Cohort Study

by Lan Huong Timm, Gasser Farrag, Martin Baxmann and Falk Schwendicke

J. Clin. Med. 2021, 10(14), 3103; https://doi.org/10.3390/jcm10143103 - 14 Jul 2021

Cited by 22 | Viewed by 4541

Abstract

Compliance is highly relevant during clear aligner therapy (CAT). In this retrospective cohort study, we assessed compliance and associated covariates in a large cohort of CAT patients. A comprehensive sample of 2644 patients (75.0% females, 25.0% males, age range 18–64 years, median 27 [...] Read more.

Compliance is highly relevant during clear aligner therapy (CAT). In this retrospective cohort study, we assessed compliance and associated covariates in a large cohort of CAT patients. A comprehensive sample of 2644 patients (75.0% females, 25.0% males, age range 18–64 years, median 27 years), all receiving CAT with PlusDental (Berlin, Germany) finished in 2019, was analyzed. Covariates included demographic ones (age, gender) as well as self-reported questionnaire-obtained ones (satisfaction with ones’ smile prior treatment, the experience of previous orthodontic therapy). The primary outcome was compliance: Based on patients’ consistent use of the mobile application for self-report and aligner wear time of ≥22 h, patients were classified as fully compliant, fairly compliant, or poorly compliant. Chi-square test was used to compare compliance in different subgroups. A total of 953/2644 (36.0%) of patients showed full compliance, 1012/2644 (38.3%) fair compliance, and 679/2644 (25.7%) poor compliance. Males were significantly more compliant than females (p = 0.000014), as were patients without previous orthodontic treatment (p = 0.023). Age and self-perceived satisfaction with ones’ smile prior to treatment were not sufficiently associated with compliance (p > 0.05). Our findings could be used to guide practitioners towards limitedly compliant individuals, allowing early intervention. Full article

► Show Figures

Figure 1

14 pages, 2459 KiB

Open AccessArticle

Association between Odontogenic and Maxillary Sinus Conditions: A Retrospective Cone-Beam Computed Tomographic Study

by Piotr Kuligowski, Aleksandra Jaroń, Olga Preuss, Ewa Gabrysz-Trybek, Joanna Bladowska and Grzegorz Trybek

J. Clin. Med. 2021, 10(13), 2849; https://doi.org/10.3390/jcm10132849 - 27 Jun 2021

Cited by 22 | Viewed by 2573

Abstract

Odontogenic infections can directly trigger maxillary sinusitis. CBCT is an excellent choice for precise examination of maxillary sinuses and hard tissues within the oral cavity. The objective of this retrospective and the cross-sectional study was to analyze the influence of odontogenic conditions on [...] Read more.

Odontogenic infections can directly trigger maxillary sinusitis. CBCT is an excellent choice for precise examination of maxillary sinuses and hard tissues within the oral cavity. The objective of this retrospective and the cross-sectional study was to analyze the influence of odontogenic conditions on the presence and intensity of maxillary sinus mucous membrane thickening using CBCT imaging. Moreover, periodontal bone loss and anatomic relationship between adjacent teeth and maxillary sinuses were assessed to evaluate its possible impact on creating maxillary thickening. The study sample consisted of 200 maxillary sinuses of 100 patients visible on CBCT examination with a field of view of 13 × 15 cm. The presented study revealed a significant influence of periapical lesions, inappropriate endodontic treatment, severe caries, and extracted teeth on the presence of increased thickening of maxillary sinus mucous membrane. In addition, an increase in the distance between root apices and maxillary sinus floor triggered a significant reduction of maxillary sinus mucous membrane thickening. The presence of periodontal bone loss significantly increases maxillary sinus mucous membrane thickening. Full article

► Show Figures

Figure 1

14 pages, 3346 KiB

Open AccessArticle

Robot-Assisted Maxillary Positioning in Orthognathic Surgery: A Feasibility and Accuracy Evaluation

by Jeong Joon Han, Sang-Yoon Woo, Won-Jin Yi and Soon Jung Hwang

J. Clin. Med. 2021, 10(12), 2596; https://doi.org/10.3390/jcm10122596 - 11 Jun 2021

Cited by 12 | Viewed by 2170

Abstract

Several methods enabling independent repositioning of the maxilla have been introduced to reduce intraoperative errors inherent in the intermediate splint. However, the accuracy is still to be improved and a different approach without time-consuming laboratory process is needed, which can allow perioperative modification [...] Read more.

Several methods enabling independent repositioning of the maxilla have been introduced to reduce intraoperative errors inherent in the intermediate splint. However, the accuracy is still to be improved and a different approach without time-consuming laboratory process is needed, which can allow perioperative modification of unoptimized maxillary position. The purpose of this study is to assess the feasibility and accuracy of a robot arm combined with intraoperative image-guided navigation in orthognathic surgery. The experiments were performed on 12 full skull phantom models. After Le Fort I osteotomy, the maxillary segment was repositioned to a different target position using a robot arm and image-guided navigation and stabilized. Using the navigation and the postoperative computed tomography (CT) images, the achieved maxillary position was compared with the planned position. Although the maxilla showed mild displacement during the fixation, the mean absolute deviations from the target position were 0.16 mm, 0.18 mm, and 0.20 mm in medio-lateral, antero-posterior, and supero-inferior directions, respectively, in the intraoperative navigation. Compared with the target position using postoperative CT, the achieved maxillary position had a mean absolute deviation of less than 0.5 mm for all dimensions and the mean root mean square deviation was 0.79 mm. The results of this study suggest that the robot arm combined with the intraoperative image-guided navigation may have great potential for surgical plan transfer with the accurate repositioning of the maxilla in the orthognathic surgery. Full article

► Show Figures

Figure 1

14 pages, 18912 KiB

Open AccessArticle

Panoptic Segmentation on Panoramic Radiographs: Deep Learning-Based Segmentation of Various Structures Including Maxillary Sinus and Mandibular Canal

by Jun-Young Cha, Hyung-In Yoon, In-Sung Yeo, Kyung-Hoe Huh and Jung-Suk Han

J. Clin. Med. 2021, 10(12), 2577; https://doi.org/10.3390/jcm10122577 - 11 Jun 2021

Cited by 21 | Viewed by 4790

Abstract

Panoramic radiographs, also known as orthopantomograms, are routinely used in most dental clinics. However, it has been difficult to develop an automated method that detects the various structures present in these radiographs. One of the main reasons for this is that structures of [...] Read more.

Panoramic radiographs, also known as orthopantomograms, are routinely used in most dental clinics. However, it has been difficult to develop an automated method that detects the various structures present in these radiographs. One of the main reasons for this is that structures of various sizes and shapes are collectively shown in the image. In order to solve this problem, the recently proposed concept of panoptic segmentation, which integrates instance segmentation and semantic segmentation, was applied to panoramic radiographs. A state-of-the-art deep neural network model designed for panoptic segmentation was trained to segment the maxillary sinus, maxilla, mandible, mandibular canal, normal teeth, treated teeth, and dental implants on panoramic radiographs. Unlike conventional semantic segmentation, each object in the tooth and implant classes was individually classified. For evaluation, the panoptic quality, segmentation quality, recognition quality, intersection over union (IoU), and instance-level IoU were calculated. The evaluation and visualization results showed that the deep learning-based artificial intelligence model can perform panoptic segmentation of images, including those of the maxillary sinus and mandibular canal, on panoramic radiographs. This automatic machine learning method might assist dental practitioners to set up treatment plans and diagnose oral and maxillofacial diseases. Full article

► Show Figures

Figure 1

13 pages, 7879 KiB

Open AccessArticle

Accurate Bracket Placement with an Indirect Bonding Method Using Digitally Designed Transfer Models Printed in Different Orientations—An In Vitro Study

by Julia Süpple, Julius von Glasenapp, Eva Hofmann, Paul-Georg Jost-Brinkmann and Petra Julia Koch

J. Clin. Med. 2021, 10(9), 2002; https://doi.org/10.3390/jcm10092002 - 07 May 2021

Cited by 14 | Viewed by 3687

Abstract

Objective: A digital workflow opens up new possibilities for the indirect bonding (IDB) of brackets. We tested if the printing orientation for bracket transfer models on the build platform of a 3D printer influences the accuracy of the following IDB method. We also [...] Read more.

Objective: A digital workflow opens up new possibilities for the indirect bonding (IDB) of brackets. We tested if the printing orientation for bracket transfer models on the build platform of a 3D printer influences the accuracy of the following IDB method. We also evaluated the clinical acceptability of the IDB method combining digitally planned and printed transfer models with the conventional fabrication of pressure-molded transfer trays. Materials and Methods: In total, 27 digitally planned bracket transfer models were printed with both 15° and 75° angulation from horizontal plane on the build platform of a digital light processing (DLP) printer. Brackets were temporarily bonded to the transfer models and pressure-molded trays were produced on them. IDB was then performed using the trays on the respective plaster models. The plaster models were scanned with an optical scanner. Digitally planned pre-bonding and scanned post-bonding bracket positions were superimposed with a software and resulted in three linear and three angular deviations per bracket. Results: No statistically significant differences of the transfer accuracy of printed transfer models angulated 15° or 75° on the 3D printer build platform were found. About 97% of the linear and 82% of the angular deviations were within the clinically acceptable range of ±0.2 mm and ±1°, respectively. The highest inaccuracies in the linear dimension occurred in the vertical towards the gingival direction and in the angular dimension in palatal crown torque. Conclusion: For the IDB method used, the printing orientation on the build platform did not have a significant impact on the transfer accuracy. Full article

► Show Figures

Figure 1

9 pages, 11861 KiB

Open AccessArticle

Impact of Image Context on Deep Learning for Classification of Teeth on Radiographs

by Joachim Krois, Lisa Schneider and Falk Schwendicke

J. Clin. Med. 2021, 10(8), 1635; https://doi.org/10.3390/jcm10081635 - 12 Apr 2021

Cited by 6 | Viewed by 2423

Abstract

Objectives: We aimed to assess the impact of image context information on the accuracy of deep learning models for tooth classification on panoramic dental radiographs. Methods: Our dataset contained 5008 panoramic radiographs with a mean number of 25.2 teeth per image. Teeth were [...] Read more.

Objectives: We aimed to assess the impact of image context information on the accuracy of deep learning models for tooth classification on panoramic dental radiographs. Methods: Our dataset contained 5008 panoramic radiographs with a mean number of 25.2 teeth per image. Teeth were segmented bounding-box-wise and classified by one expert; this was validated by another expert. Tooth segments were cropped allowing for different context; the baseline size was 100% of each box and was scaled up to capture 150%, 200%, 250% and 300% to increase context. On each of the five generated datasets, ResNet-34 classification models were trained using the Adam optimizer with a learning rate of 0.001 over 25 epochs with a batch size of 16. A total of 20% of the data was used for testing; in subgroup analyses, models were tested only on specific tooth types. Feature visualization using gradient-weighted class activation mapping (Grad-CAM) was employed to visualize salient areas. Results: F1-scores increased monotonically from 0.77 in the base-case (100%) to 0.93 on the largest segments (300%; p = 0.0083; Mann–Kendall-test). Gains in accuracy were limited between 200% and 300%. This behavior was found for all tooth types except canines, where accuracy was much higher even for smaller segments and increasing context yielded only minimal gains. With increasing context salient areas were more widely distributed over each segment; at maximum segment size, the models assessed minimum 3–4 teeth as well as the interdental or inter-arch space to come to a classification. Conclusions: Context matters; classification accuracy increased significantly with increasing context. Full article

► Show Figures

Figure 1

13 pages, 1557 KiB

Open AccessArticle

Barriers and Enablers for Artificial Intelligence in Dental Diagnostics: A Qualitative Study

by Anne Müller, Sarah Marie Mertens, Gerd Göstemeyer, Joachim Krois and Falk Schwendicke

J. Clin. Med. 2021, 10(8), 1612; https://doi.org/10.3390/jcm10081612 - 10 Apr 2021

Cited by 18 | Viewed by 3934

Abstract

The present study aimed to identify barriers and enablers for the implementation of artificial intelligence (AI) in dental, specifically radiographic, diagnostics. Semi-structured phone interviews with dentists and patients were conducted between the end of May and the end of June 2020 (convenience/snowball sampling). [...] Read more.

The present study aimed to identify barriers and enablers for the implementation of artificial intelligence (AI) in dental, specifically radiographic, diagnostics. Semi-structured phone interviews with dentists and patients were conducted between the end of May and the end of June 2020 (convenience/snowball sampling). A questionnaire developed along the Theoretical Domains Framework (TDF) and the Capabilities, Opportunities and Motivations influencing Behaviors model (COM-B) was used to guide interviews. Mayring’s content analysis was employed to point out barriers and enablers. We identified 36 barriers, conflicting themes or enablers, covering nine of the fourteen domains of the TDF and all three determinants of behavior (COM). Both stakeholders emphasized chances and hopes for AI. A range of enablers for implementing AI in dental diagnostics were identified (e.g., the chance for higher diagnostic accuracy, a reduced workload, more comprehensive reporting and better patient–provider communication). Barriers related to reliance on AI and responsibility for medical decisions, as well as the explainability of AI and the related option to de-bug AI applications, emerged. Decision-makers and industry may want to consider these aspects to foster implementation of AI in dentistry. Full article

► Show Figures

Figure 1

12 pages, 3105 KiB

Open AccessArticle

A Validation Employing Convolutional Neural Network for the Radiographic Detection of Absence or Presence of Teeth

by María Prados-Privado, Javier García Villalón, Antonio Blázquez Torres, Carlos Hugo Martínez-Martínez and Carlos Ivorra

J. Clin. Med. 2021, 10(6), 1186; https://doi.org/10.3390/jcm10061186 - 12 Mar 2021

Cited by 2 | Viewed by 2523

Abstract

Dental radiography plays an important role in clinical diagnosis, treatment and making decisions. In recent years, efforts have been made on developing techniques to detect objects in images. The aim of this study was to detect the absence or presence of teeth using [...] Read more.

Dental radiography plays an important role in clinical diagnosis, treatment and making decisions. In recent years, efforts have been made on developing techniques to detect objects in images. The aim of this study was to detect the absence or presence of teeth using an effective convolutional neural network, which reduces calculation times and has success rates greater than 95%. A total of 8000 dental panoramic images were collected. Each image and each tooth was categorized, independently and manually, by two experts with more than three years of experience in general dentistry. The neural network used consists of two main layers: object detection and classification, which is the support of the previous one. A Matterport Mask RCNN was employed in the object detection. A ResNet (Atrous Convolution) was employed in the classification layer. The neural model achieved a total loss of 0.76% (accuracy of 99.24%). The architecture used in the present study returned an almost perfect accuracy in detecting teeth on images from different devices and different pathologies and ages. Full article

► Show Figures

Figure 1

18 pages, 3790 KiB

Open AccessArticle

Accuracy, Labor-Time and Patient-Reported Outcomes with Partially versus Fully Digital Workflow for Flapless Guided Dental Implants Insertion—A Randomized Clinical Trial with One-Year Follow-Up

by Corina Marilena Cristache, Mihai Burlibasa, Ioana Tudor, Eugenia Eftimie Totu, Fabrizio Di Francesco and Liliana Moraru

J. Clin. Med. 2021, 10(5), 1102; https://doi.org/10.3390/jcm10051102 - 06 Mar 2021

Cited by 12 | Viewed by 2731

Abstract

(1) Background: Prosthetically-driven implant positioning is a prerequisite for long-term successful treatment. Transferring the planned implant position information to the clinical setting could be done using either static or dynamic guided techniques. The 3D model of the bone and surrounding structures is obtained [...] Read more.

(1) Background: Prosthetically-driven implant positioning is a prerequisite for long-term successful treatment. Transferring the planned implant position information to the clinical setting could be done using either static or dynamic guided techniques. The 3D model of the bone and surrounding structures is obtained via cone beam computed tomography (CBCT) and the patient’s oral condition can be acquired conventionally and then digitalized using a desktop scanner, partially digital workflow (PDW) or digitally with the aid of an intraoral scanner (FDW). The aim of the present randomized clinical trial (RCT) was to compare the accuracy of flapless dental implants insertion in partially edentulous patients with a static surgical template obtained through PDW and FDW. Patient outcome and time spent from data collection to template manufacturing were also compared. (2) Methods: 66 partially edentulous sites (at 49 patients) were randomly assigned to a PDW or FDW for guided implant insertion. Planned and placed implants position were compared by assessing four deviation parameters: 3D error at the entry point, 3D error at the apex, angular deviation, and vertical deviation at entry point. (3) Results: A total of 111 implants were inserted. No implant loss during osseointegration or mechanical and technical complications occurred during the first-year post-implants loading. The mean error at the entry point was 0.44 mm (FDW) and 0.85 (PDW), p ≤ 0.00; at implant apex, 1.03 (FDW) and 1.48 (PDW), p ≤ 0.00; the mean angular deviation, 2.12° (FDW) and 2.48° (PDW), p = 0.03 and the mean depth deviation, 0.45 mm (FDW) and 0.68 mm (PDW), p ≤ 0.00; (4) Conclusions: Despite the statistically significant differences between the groups, and in the limits of the present study, full digital workflow as well as partially digital workflow are predictable methods for accurate prosthetically driven guided implants insertion. Full article

► Show Figures

Figure 1

8 pages, 1019 KiB

Open AccessArticle

Generalizability of Deep Learning Models for Caries Detection in Near-Infrared Light Transillumination Images

by Agnes Holtkamp, Karim Elhennawy, José E. Cejudo Grano de Oro, Joachim Krois, Sebastian Paris and Falk Schwendicke

J. Clin. Med. 2021, 10(5), 961; https://doi.org/10.3390/jcm10050961 - 01 Mar 2021

Cited by 19 | Viewed by 2962

Abstract

Objectives: The present study aimed to train deep convolutional neural networks (CNNs) to detect caries lesions on Near-Infrared Light Transillumination (NILT) imagery obtained either in vitro or in vivo and to assess the models’ generalizability. Methods: In vitro, 226 extracted posterior permanent human [...] Read more.

Objectives: The present study aimed to train deep convolutional neural networks (CNNs) to detect caries lesions on Near-Infrared Light Transillumination (NILT) imagery obtained either in vitro or in vivo and to assess the models’ generalizability. Methods: In vitro, 226 extracted posterior permanent human teeth were mounted in a diagnostic model in a dummy head. Then, NILT images were generated (DIAGNOcam, KaVo, Biberach), and images were segmented tooth-wise. In vivo, 1319 teeth from 56 patients were obtained and segmented similarly. Proximal caries lesions were annotated pixel-wise by three experienced dentists, reviewed by a fourth dentist, and then transformed into binary labels. We trained ResNet classification models on both in vivo and in vitro datasets and used 10-fold cross-validation for estimating the performance and generalizability of the models. We used GradCAM to increase explainability. Results: The tooth-level prevalence of caries lesions was 41% in vitro and 49% in vivo, respectively. Models trained and tested on in vivo data performed significantly better (mean ± SD accuracy: 0.78 ± 0.04) than those trained and tested on in vitro data (accuracy: 0.64 ± 0.15; p < 0.05). When tested in vitro, the models trained in vivo showed significantly lower accuracy (0.70 ± 0.01; p < 0.01). Similarly, when tested in vivo, models trained in vitro showed significantly lower accuracy (0.61 ± 0.04; p < 0.05). In both cases, this was due to decreases in sensitivity (by −27% for models trained in vivo and −10% for models trained in vitro). Conclusions: Using in vitro setups for generating NILT imagery and training CNNs comes with low accuracy and generalizability. Clinical significance: Studies employing in vitro imagery for developing deep learning models should be critically appraised for their generalizability. Applicable deep learning models for assessing NILT imagery should be trained on in vivo data. Full article

► Show Figures

Figure 1

2020

Jump to: 2023, 2022, 2021

2 pages, 149 KiB

Open AccessEditorial

Digital Dentistry: Advances and Challenges

by Falk Schwendicke

J. Clin. Med. 2020, 9(12), 4005; https://doi.org/10.3390/jcm9124005 - 11 Dec 2020

Cited by 6 | Viewed by 2015

Abstract

Dental diseases like caries or periodontitis are among the most prevalent in the world, generating significant subjective and financial burden to individuals and healthcare systems [...] Full article

19 pages, 1840 KiB

Open AccessArticle

Clinical Outcomes of Root-Analogue Implants Restored with Single Crowns or Fixed Dental Prostheses: A Retrospective Case Series

by Mats Wernfried Heinrich Böse, Detlef Hildebrand, Florian Beuer, Christian Wesemann, Paul Schwerdtner, Stefano Pieralli and Benedikt Christopher Spies

J. Clin. Med. 2020, 9(8), 2346; https://doi.org/10.3390/jcm9082346 - 23 Jul 2020

Cited by 11 | Viewed by 2946

Abstract

The objective was to investigate clinical and radiological outcomes of rehabilitations with root-analogue implants (RAIs). Patients restored with RAIs, supporting single crowns or fixed dental prostheses, were recruited for follow-up examinations. Besides clinical and esthetical evaluations, X-rays were taken and compared with the [...] Read more.

The objective was to investigate clinical and radiological outcomes of rehabilitations with root-analogue implants (RAIs). Patients restored with RAIs, supporting single crowns or fixed dental prostheses, were recruited for follow-up examinations. Besides clinical and esthetical evaluations, X-rays were taken and compared with the records. Patients were asked to evaluate the treatment using Visual Analogue Scales (VAS). For statistical analyses, mixed linear models were used. A total of 107 RAIs were installed in one dental office. Of these, 31 were available for follow-up examinations. For those remaining, survival has been verified via phone. RAIs were loaded after a mean healing time of 6.6 ± 2.5 months. 12.1 ± 6.9 months after loading, a mean marginal bone loss (MBL) of 1.20 ± 0.73 mm was measured. Progression of MBL significantly decreased after loading (p = 0.013). The mean pink and white esthetic score (PES/WES) was 15.35 ± 2.33 at follow-up. A survival rate of 94.4% was calculated after a mean follow-up of 18.9 ± 2.4 months after surgery. Immediate installation of RAIs does not seem to reduce MBL, as known from the literature regarding screw-type implants, and might not be recommended for daily routine. Nevertheless, they deliver esthetically satisfying results. Full article

► Show Figures

Figure 1

12 pages, 1174 KiB

Open AccessArticle

How Accurate Is Oral Implant Installation Using Surgical Guides Printed from a Degradable and Steam-Sterilized Biopolymer?

by Stefano Pieralli, Benedikt Christopher Spies, Valentin Hromadnik, Robert Nicic, Florian Beuer and Christian Wesemann

J. Clin. Med. 2020, 9(8), 2322; https://doi.org/10.3390/jcm9082322 - 22 Jul 2020

Cited by 33 | Viewed by 3299

Abstract

3D printed surgical guides are used for prosthetically-driven oral implant placement. When manufacturing these guides, information regarding suitable printing techniques and materials as well as the necessity for additional, non-printed stock parts such as metal sleeves is scarce. The aim of the investigation [...] Read more.

3D printed surgical guides are used for prosthetically-driven oral implant placement. When manufacturing these guides, information regarding suitable printing techniques and materials as well as the necessity for additional, non-printed stock parts such as metal sleeves is scarce. The aim of the investigation was to determine the accuracy of a surgical workflow for oral implant placement using guides manufactured by means of fused deposition modeling (FDM) from a biodegradable and sterilizable biopolymer filament. Furthermore, the potential benefit of metal sleeve inserts should be assessed. A surgical guide was designed for the installation of two implants in the region of the second premolar (SP) and second molar (SM) in a mandibular typodont model. For two additive manufacturing techniques (stereolithography [SLA]: reference group, FDM: observational group) n = 10 surgical guides, with (S) and without (NS) metal sleeves, were used. This resulted in 4 groups of 10 samples each (SLA-S/NS, FDM-S/NS). Target and real implant positions were superimposed and compared using a dedicated software. Sagittal, transversal, and vertical discrepancies at the level of the implant shoulder, apex and regarding the main axis were determined. MANOVA with posthoc Tukey tests were performed for statistical analyses. Placed implants showed sagittal and transversal discrepancies of <1 mm, vertical discrepancies of <0.6 mm, and axial deviations of ≤3°. In the vertical dimension, no differences between the four groups were measured (p ≤ 0.054). In the sagittal dimension, SLA groups showed decreased deviations in the implant shoulder region compared to FDM (p ≤ 0.033), whereas no differences in the transversal dimension between the groups were measured (p ≤ 0.054). The use of metal sleeves did not affect axial, vertical, and sagittal accuracy, but resulted in increased transversal deviations (p = 0.001). Regarding accuracy, biopolymer-based surgical guides manufactured by means of FDM present similar accuracy than SLA. Cytotoxicity tests are necessary to confirm their biocompatibility in the oral environment. Full article

► Show Figures

Figure 1

Journal Menu

Journal Browser

Digital Dentistry: Advances and Challenges (Closed)

Share This Topical Collection

Editor

Topical Collection Information

Keywords

Published Papers (22 papers)

2023

Jump to: 2022, 2021, 2020

2022

Jump to: 2023, 2021, 2020

2021

Jump to: 2023, 2022, 2020

2020

Jump to: 2023, 2022, 2021

Further Information

Guidelines

MDPI Initiatives

Follow MDPI