Artificial intelligence represented by deep learning has attracted attention in the field of dental implant restoration. It is widely used in surgical image analysis, implant plan design, prosthesis shape design, and prognosis judgment. This article mainly describes the research progress of deep learning in the whole process of dental implant prosthodontics. It analyzes the limitations of current research, and looks forward to the future development direction.
There is evidence that the apico-coronal implant position and the mucosal phenotype can affect the extent of peri-implant bone loss. This clinical trial analyzes the bone remodeling and marginal bone loss that occur around conical-connection implants placed equicrestally and subcrestally, assessing the effect of the peri-implant soft-tissue phenotype.
Fifty-one patients received 56 implants of distinct diameters (3.5 mm Ø n = 6; 4.3 mm Ø n = 41; 5 mm Ø n = 9) in the posterior part of the maxilla or mandible. The implants were placed equicrestally, 1 mm subcrestally and >1 mm subcrestally, depending on the initial supracrestal tissue height (STH). After 3 months of non-submerged healing, single metal-ceramic screw-retained implant-supported crowns were placed. Longitudinal measurements of STH, mucosal thickness and keratinized mucosa width (KMW) were made at the time of implant placement (T0), crown placement (T1), and after 3 (T2) and 6 months (T3) of prosthetic loading. At each of these points, a radiographic evaluation of bone remodeling and marginal bone loss was also performed.
STH was significantly greater for implants placed >1 mm subcrestally than for those placed 1 mm subcrestally. After 12 months of follow-up, a very significant (p < 0.001) loss of KMW was observed, in addition to a marginal bone loss of 0.08 ± 0.1, 0.15 ± 0.2, and 0.14 ± 0.2 mm in the groups placed equicrestally, 1 mm subcrestally and >1 mm subcrestally, respectively. After the multiple linear regression, marginal bone loss was found to depend primarily on KMW (β = −0.43), while also being affected by STH (β = 0.32) and implant diameter (β = −0.28).
Marginal bone loss may be influenced by the position with respect to the bone crest, as well as the KMW, STH, and implant diameter. However, more well-controlled studies are needed to verify these above-mentioned findings with different implant designs and connections.
Oro-control communication is one of the complications associated with dental extraction and oral surgeries. This case report presents a minimally invasive surgical approach for bone regeneration at the site of oro-antral communication utilizing a prefabricated computer-aided design and computer-aided manufacturing (CAD-CAM) allogenic bone block.
A 20-year-old healthy female, nonsmoker, with a badly destructed upper right first molar was referred for dental implant placement after extraction. Cone beam computerized tomography images revealed the presence of a large bone defect associated with oro-antral communication with the maxillary sinus and insufficient bone for dental implant placement. A prefabricated CAD-CAM allogenic bone scaffold was fabricated. After surgical exposure, the scaffold was secured in place and covered with a non-resorbable membrane. A dental implant was placed after 5 months, and a trephining biopsy was processed for histological evaluation.
Closure of the oro-antral communication was clinically observed. The average width of the alveolar bone was 12 mm, and the average height was 11 mm. Histological analysis at 5-month intervals showed thin newly formed bone trabeculae encircling remnants of graft material surrounded by osteoid tissue. The newly formed bone percentages were 32 ± 18% and 28 ± 17% volume remained after the biodegradation of the scaffold. Specific immune-histochemical staining by anti-vascular epithelial growth factor expression index value was 32.06%.
A prefabricated CAD-CAM scaffold was successfully used to seal a large oro-antral communication and regenerate sufficient bone to place a dental implant.
To assess the accuracy of autonomous robotic and fully guided static computer-assisted implant surgery (sCAIS) performed on models and patients.
This study was divided into in vitro and in vivo sections. In vitro, 80 operators were assigned to two groups randomly. Forty operators performed forty autonomous robotic implant (ARI group) surgeries and the remaining forty operators carried out forty fully guided sCAIS (FGI group) surgeries on maxillary models, respectively. Each operator placed an implant in one maxillary model. In vivo, 60 patients with 113 implants from 2019 to 2023 (ARI group: 32 patients, 58 implants; FGI group: 28 patients, 55 implants) receiving implant surgeries were incorporated in this retrospective research. The preoperative and postoperative cone beam computer tomographs (CBCTs) were utilized to estimate the linear deviations and angular deviations in two-dimensional (2D) and three-dimensional (3D) space. The Pearson's chi-square test, Shapiro–Wilk test, Student's t test, Mann–Whitney U test and mixed models were applied, and p <0.05 was considered statistically significant.
In vitro, a total of 80 implants were enrolled and significant differences were found between the two groups (p < 0.001): The 3D deviation at the platform of ARI and FGI group was 0.58 ± 0.60 mm and 1.50 ± 1.46 mm, respectively, at the apex was 0.58 ± 0.60 mm and 1.78 ± 1.35 mm, respectively, and angle was 1.01 ± 0.87° and 2.93 ± 1.59°, respectively. Also, except for mesiodistal deviation at the implant platform, the rest linear and angular deviations in the ARI group were significantly lower than those in the FGI group in 2D space (p < 0.001). In vivo, a significantly lower mean of angular deviation (0.95 ± 0.50°, p < 0.001) and the linear deviation at both platform (0.45 ± 0.28 mm, p < 0.001) and apex (0.47 ± 0.28 mm, p < 0.001) were observed in ARI group when compared to the FGI group (4.31 ± 2.60°; 1.45 ± 1.27 mm; 1.77 ± 1.14 mm).
The use of autonomous robotic technology showed significantly higher accuracy than the fully guided sCAIS.
To assess and compare the short-term clinical outcomes following a combined vertical and horizontal alveolar ridge augmentation and two-stage implant placement using either autogenous tooth roots (TR) or autogenous bone blocks (AB).
A total of n = 27 patients (TR/AB: 13/14) exhibiting n = 31 implants (TR/AB: 14/17) were available for the analysis. Each subject had been allocated to a combined vertical and horizontal alveolar ridge augmentation using either (1) healthy TR (e.g., retained wisdom teeth), or (2) monocortical AB harvested from the retromolar area (i.e. external oblique line). Clinical parameters (e.g., bleeding on probing, BOP; probing pocket depth, PD; mucosal recession, MR) were recorded after a follow-up period of 16.03 ± 4.3 months following implant placement.
The survival rates amounted to 100% in both groups. TR and AB grafted sites were associated with similar changes in mean BOP (8.97 ± 27.73%; 11.90 ± 18.97%), PD (0.53 ± 0.49; 0.47 ± 0.67 mm), and MR (0.03 ± 0.13; 0.0 ± 0.02 mm) values. The incidence of peri-implant mucositis and peri-implantitis at the patient level amounted to 15.38% and 0.0% in the TR-, and 28.57% and 7.14% in the AB group.
Both surgical procedures were associated with peri-implant tissue health and stability on the short-term.
This study aimed to use a deep learning (DL) approach for the automatic identification of the ridge deficiency around dental implants based on an image slice from cone-beam computerized tomography (CBCT).
Single slices crossing the central long-axis of 630 mandibular and 845 maxillary virtually placed implants (4–5 mm diameter, 10 mm length) in 412 patients were used. The ridges were classified based on the intraoral bone-implant support and sinus floor location. The slices were either preprocessed by alveolar ridge homogenizing prior to DL (preprocessed) or left unpreprocessed. A convolutional neural network with ResNet-50 architecture was employed for DL.
The model achieved an accuracy of >98.5% on the unpreprocessed image slices and was found to be superior to the accuracy observed on the preprocessed slices. On the mandible, model accuracy was 98.91 ± 1.45%, and F1 score, a measure of a model's accuracy in binary classification tasks, was lowest (97.30%) on the ridge with a combined horizontal-vertical defect. On the maxilla, model accuracy was 98.82 ± 1.11%, and the ridge presenting an implant collar-sinus floor distance of 5–10 mm with a dehiscence defect had the lowest F1 score (95.86%). To achieve >90% model accuracy, ≥441 mandibular slices or ≥592 maxillary slices were required.
The ridge deficiency around dental implants can be identified using DL from CBCT image slices without the need for preprocessed homogenization. The model will be further strengthened by implementing more clinical expertise in dental implant treatment planning and incorporating multiple slices to classify 3-dimensional implant-ridge relationships.
The primary aim of this systematic review was to investigate and compare the outcomes of different vertical ridge augmentation (VRA) techniques in relation to peri-implant bone loss (PBL), after at least 12 months of functional loading.
The search was conducted to find all the studies about VRA and measurements of PBL with at least 12 months follow-up. Three pairwise meta-analysis (MA) was performed to completely evaluate the outcomes.
A total of 42 studies were included, of which 11 were randomized clinical trials (RCTs). RCTs were available only for guided bone regeneration (GBR), onlay, and inlay techniques. The weighted mean estimate (WME) of PBL value was found to be 1.38 mm (95% confidence interval [95% CI]: 1.10–1.66) after a mean follow-up of 41.0 ± 27.8 months. GBR, Inlay, Onlay, osteodistraction, and SBB represented in weight 32.9%, 30.6%, 25.0%, 7.6%, and 3.9%, respectively; and their WME (95% CI) were 1.06 (0.87–1.26) mm, 1.72 (1.00–2.43) mm, 1.31 (0.87–1.75) mm, 1.81 (0.87–1.75) mm, and 0.66 (0.55–0.77) mm, respectively. Among the secondary outcomes, the analysis was conducted for vertical bone gain, healing complication rate, surgical complication rate, implant survival, and success rate.
The primary findings of the meta-analysis, based on the changes between final and baseline values, showed that the peri-implant bone loss could be influenced by the type of intervention but there is a need to evaluate in RCTs the behavior of the peri-implant bone levels after long-term follow-up for all techniques.
To assess both the radiographic and profilometric outcomes of early implant placement with or without alveolar ridge preservation (ARP) (using two different ARP techniques) after 1 year of loading.
Seventy-five patients with a failing single tooth in the anterior maxilla were randomly allocated to three groups (1:1:1): (a) ARP using demineralized bovine bone mineral containing 10% collagen (DBBM-C) covered by a collagen matrix (CM), (b) ARP using DBBM-C covered with a palatal graft (PG), and (c) unassisted socket healing (control). Eight weeks after tooth extraction, early implant placement was performed in all patients. Cone-beam computed tomography (CBCT) and impressions were taken 8 weeks after tooth extraction (ARP/unassisted healing) prior to implant placement and 1-year post-loading. Radiographic and profilometric outcomes were evaluated.
Out of the 70 patients available for re-examination at 1-year post-loading, 55 datasets could be assessed (ARP-CM 19; ARP-PG 17; Control 19). The need for additional guided bone regeneration (GBR) at implant placement amounted to 31.6% (ARP-CM), 29.4% (ARP-PG), and 68.4% (unassisted healing). Adjusted models revealed that residual buccal bone height and additional GBR at implant placement significantly influenced the magnitude of the alveolar changes at 1 year (p < 0.05). In patients with ARP (group ARP-CM or ARP-PG) without additional GBR, the presence of bone convexity amounted to 36.0% (9/25) at 1-year post-loading. For patients that received ARP and additional GBR at implant placement, the frequency of bone convexity increased to 72.7% (8/11) (p = 0.042). Regarding profilometric measurements, a tendency toward agreement with radiographic outcomes was observed.
Early implant placement with ARP can attenuate alveolar ridge changes at 1-year post loading by minimizing both radiographic and profilometric alterations. However, early implant placement with simultaneous GBR consistently yields superior radiographic and profilometric outcomes, regardless of whether ARP is performed.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：