For the success of dental implants, the implant surface must be covered with bone. There is a positive correlation between alveolar bone thickness and primary osteointegration(1,2). Peri-implant defects are supportive bone tissue loss with a prevalence ranging from 28% to 56%(3). Routine radiographic evaluations are a technique used to evaluate whether peri-implantitis develops. When peri-implantitis goes unnoticed early on, marginal bone loss progresses and leaves the clinician with increasingly narrowing treatment options. From the patient’s point of view, there will be a decrease in the quality of final oral rehabilitation unless the condition is promptly addressed(4).
Intraoral and panoramic radiography techniques show the mesial and distal areas of the bone(5); but there can be geometric distortions and anatomical superimpositions(6). If 3-dimensional observation of the bone is necessary, conventional and cone-beam computed tomography (CBCT) can provide alternative options(7). There are certain major limitations of these techniques, though, including their high cost, increased radiation exposure, and formation of metal artifacts(8).
In view of such disadvantages of CBCT, the effectiveness of ultrasonography (US) on bone surface evaluation, bone thickness measurement, and peri-implant defects visibility has started to be evaluated in the literature in recent years(9–12). It is reported that US is the preferred modality due to its advantages such as non-invasive nature, use of non-ionizing radiation, good tolerability by patients, and low cost(13,14).
US is based on the principle of measuring the energy loss caused by the emission, reflection, and scattering of acoustic waves lower than 20 kHz in different tissues. The energy loss of the wave propagating throughout the tissue is associated with the acoustic properties of the waves(15). Due to the high-frequency mode on US, the depth of signal penetration into the tissue decreases, but the image quality is improved. This means that there is an inverse correlation between the image resolution and the measured depth. It has been reported in the literature that high-frequency ultrasounds can be used to scan the bone surface(16). In another study, it was found that the combined use of high- and low-frequency ultrasound may be a new approach in cortical bone evaluation(9).
The present study aimed to evaluate the efficacy of intraoral and extraoral US evaluations performed with two different types of probes (linear and hockey stick) for the visibility of peri-implant bone defects.
This study was carried out using three sheep heads including soft tissues with 14 implants [zirconium (
Two US consoles (ProSound Alpha 6, Hitachi Aloka Medical Ltd., Tokyo, Japan) with a hockey stick intra-operative probe, 13 MHz (UST-536, Hitachi Aloka Medical Ltd., Tokyo, Japan), and a linear probe, 5–13.3 MHz (UST-5413, Hitachi Aloka Medical Ltd., Tokyo, Japan), and also a high-resolution ACUSONS 2000 ultrasound unit (Siemens, Munich, Germany) with a 4–9 MHz linear probe (9L4 Transducer) and a hockey stick intra-operative probe 14 MHz (14L5 SP Transducer) were used in the study. All fenestrations and dental implants were scanned both intraorally and extraorally (Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5). Extraoral scanning was performed with a linear probe, while intraoral scanning was done only with hockey stick probes with a different frequency bandwidth. After adjusting the probe to the desired trajectory, US images were obtained using both probes. Two observers (H.K., K.O) carried out all US scanning and evaluations. The observers were dentomaxillofacial radiologists with 10 years’ and 18 years’ of experience with US, respectively.
The two observers conducted two separate US sessions independently. The study was performed twice, with an interval of 2 weeks after the initial US scanning. The same sheep heads were used for both US scanning procedures. The observers were free to position the probe when taking images.
Before the US evaluations, both observers were trained to appropriately use the US software in a special session. However, no calibration was made for the US evaluations since the scanning itself was in real-time. The fenestration defects were detected simultaneously in each US session during scanning, so the detection of defects was done in real-time while scanning. US system proprietary software was used (Hitachi Aloka Medical Ltd., Tokyo, Japan, Siemens S2000, Munich, Germany). The observers were free to use any enhancement procedure that was available in the US unit. Moreover, the observers were aware of the existence of the defects, however they did not know the dental implant type (titanium or zirconium).
For all imaging methods, a five-point scale was used to assess the visibility of each fenestration: (1) definitely absent; (2) probably absent; (3) unsure; (4) probably present; (5) definitely present.
Intraclass Correlation (ICC) analysis was used for the assessment of intraobserver and intraobserver reliability. Kappa coefficients were calculated to evaluate the gold standard and the observers’ agreements for each image set. Kappa values were interpreted according to the guidelines proposed by Landis and Koch(17), and adapted by Altman(18) κ ≤0.20, poor; κ = 0.21–0.40, fair; κ = 0.41–0.60, moderate; κ = 0.61–0.80, good; and κ = 0.81–1.00, very good. Scores obtained from the (1) 1st linear probe extraorally; (2) 2nd linear probe extraorally; (3) 1st hockey stick probe intraorally; and (4) 2nd hockey stick probe intraorally were compared with the gold standard. A probability level of less than 5% (
The lowest ICC value in the intraobserver reliability assessment was obtained with a linear probe (UST-5413) that was used extraorally. The highest ICC value in the intraobserver reliability assessment was obtained with a 14 MHz (14L5 SP Transducer) hockey stick probe (0.966) for Observer 1, and with a 13 MHz (UST-536 Transducer) hockey stick probe that was used intraorally (0.952) for Observer 2 (Tab. 1).
Intraobserver agreement for Observer 1 and Observer 2. p value less than 0.05 considered as statically significant (95% CI)
Observer 1 | Observer 2 | |||
---|---|---|---|---|
ICC |
|
ICC |
|
|
|
0.878 |
0.0001 | 0.952 |
0.001 |
|
0.966 |
0.0001 | 0.921 |
0.001 |
|
0.696 |
0.019 | 0.727 |
0.016 |
|
0.849 |
0.001 | 0.875 |
0.001 |
The interobserver ICC coefficients were presented in Tab. 2. Good interobserver reliability was achieved in all probes with ICC values between 0.762 and 0.914. The linear probe (UST-5413) that was used extraorally had the lowest interobserver reliability (Tab. 2).
Interobserver agreement between observers. p value less than 0.05 considered as statically significant (95% CI)
ICC |
|
|
---|---|---|
|
0.784 (0.349-0.930) | 0.002 |
|
0.909 (0.721-0.971) | 0.001 |
|
0.762 (0.300-0.922) | 0.006 |
|
0.914 (0.728-0.972) | 0.001 |
The 13 MHz hockey stick intraoral probe (UST-536 Transducer) had a high level of agreement with the gold standard (
Comparison of observers with the gold standard. p value less than 0.05 considered as statically significant
Sensitivity | Specificity | PPV | NPV | κ |
|
||
---|---|---|---|---|---|---|---|
|
Observer 1-Gold Standard | 80% | 100% | 100% | 90% | 0.837 | 0.001 |
|
Observer 1-Gold Standard | 80% | 89% | 80% | 89%hz | 0.689 | 0.011 |
|
Observer 1-Gold Standard | 100% | 33% | 46% | 100% | 0.263 | 0.145 |
|
Observer 1-Gold Standard | 100% | 33% | 46% | 100% | 0.263 | 0.145 |
|
Observer 2-Gold Standard | 100% | 89% | 83% | 100% | 0.851 | 0.001 |
|
Observer 2-Gold Standard | 100% | 78% | 71% | 100% | 0.714 | 0.005 |
|
Observer 2-Gold Standard | 60% | 22% | 30% | 50% | –0.145 | 0.481 |
|
Observer 2-Gold Standard | 80% | 11% | 33% | 50% | –0.068 | 0.649 |
PPV – positive predictive value, NPV – negative predictive value
The agreement between Observer 2 with the gold standard was high (
Early diagnostic criteria of peri-implantitis include radiographic bone loss greater than one-third of implant height(19). For this reason, it is important to monitor bone loss in the follow-up period after the operation(20). In dentistry, radiographic examination is the most common choice to evaluate peri-implantitis(5). In the meta-analysis of Bohner
To the best of our knowledge, there are only a few studies on bone surfaces with evaluations performed using US(9,13,16,24). In their study, Degen
Several studies have explored the effect of probes of different frequencies on the quality of imaging periodontal defects. Mahmoud
CBCT can also be used routinely for detecting these kinds of defects. However, the occurrence of metal artifacts around dental implants, as scattering or complete absorption of the beam can exist and be concluded with image degradation. This situation can prevent the observation of the implant-bone interface, and make it difficult to evaluate peri-implant bone defects(27,28). In the head and neck area, high-resolution images in multiple planes are obtained with modern US units with high-frequency linear probes (7.5–12 MHz)(29). In dental practice, US is mainly used in cases of maxillofacial fractures(30), cervical lymphadenopathy(31),soft tissue masses(32), masticatory and neck muscles(33,34), temporomandibular joint(35), periapical(36,37) salivary gland diseases(38), intraosseous jaw pathologies(39), and carotid paragangliomas(40).
There are several limitations of this study. Firstly, two different ultrasound systems with probes slightly differing in frequency were used. Although the frequencies of probes are similar to each other, they may still influence the results obtained. This can be due to the value of the signal to noise ratio (SNR) which increases as the frequency of the ultrasound signal rises. It was stated that less speckle noise was produced as higher frequency signal was applied(41). In a recent paper, it was also found that lower frequency may achieve a better depth penetration, while higher frequencies are associated with better resolution. While the output power may improve image quality by increasing the intensity of transmitted sound energy, the impact is usually insignificant(42). This issue can be addressed in more depth in future studies.
Another limitation of this study is that even though two observers performed two separate US sessions independently, observational differences in US can affect the results. Also, no comparison was performed for implant types (zirconia and titanium implants). Since the number of implants of both types in this study is not sufficient, the comparison between them has not been studied statistically. However, further studies with more implant types and numbers should be done to elucidate the differences between implant types in US images.
Hockey stick probes used intraorally can be an effective option for the evaluation of the visibility of peri-implant bone defects. No report was found in the literature regarding a comparison of peri-implant bone defect visibility with different US probes. To the best of our knowledge, this study is the first one addressing this subject. Thus, the conclusion is that US can be an alternative method of evaluating defects. However, further studies are needed to determine the effectiveness of US in the visualization of peri-implant bone defects.