*Case Report* **Direct Visualization of Cervical Interlaminar Epidural Injections Using Sonography**

**Nana Maeda 1, Manabu Maeda 1,\* and Yasuhito Tanaka <sup>2</sup>**


**Abstract:** In this case series, we describe a novel ultrasound (US)-guided cervical interlaminar epidural steroid injections (CILESIs) procedure that does not depend on the loss-of-resistance method for epidural space identification. A needle is introduced into three US-identified structures (triple bar sign), the interspinal ligament, ligamentum flavum, and dura mater. The injectants are monitored using superb microvascular imaging during injection. Here, we demonstrate the use of US-guided CILESIs in nine cases and propose the use of sonography, rather than conventional methods, for easier and safer cervical epidural injections. Sonography for direct visualization of cervical epidural injection may allow for outpatient injections.

**Keywords:** ultrasound; epidural injection; cervical; steroid; superb microvascular imaging

#### **1. Introduction**

A major concern for cervical epidural injection (CEDI) in patients with neck pain and cervical radiculopathy is the safety and effectiveness of the procedure. Reports of serious complications such as spinal cord injury have led to concerns regarding this procedure's safety [1–3]. Fluoroscopy or computed tomography (CT) guidance aims to reduce the risks of epidural injection, such as dural puncture or spinal cord injury. However, although both guidance approaches improve safety, they cannot completely aid in advancing the tip of the needle into the epidural space, as these approaches cannot visualize the entire intrusion route [2–5].

It is also important to examine the placement of the needle tip and the spread of the injectant to ensure its effectiveness [6]. The loss-of-resistance (LOR) method is performed for blinded or fluoroscopic epidural injections. However, it is limited by false-positive LOR before the needle enters the epidural space [7,8]. CT guidance uses high radiation exposure to carefully check the needle tip position, which also poses a problem [9]. However, CT cannot image the epidural space without a contrast medium, which can cause adverse effects [10]. Furthermore, CT emits ionizing radiation, which can cause cancer [11].

Sonography offers an alternative method for performing cervical interlaminar epidural steroid injections (CILESIs). It could help reduce the risk of injury, avoid the pitfall of falsepositive LOR, and avoid radiation exposure or the use of contrast media by allowing direct visualization of the location of the needle tip during the procedure with cross-sectional imaging [6].

However, the technique for ultrasound (US)-guided CILESIs has not been reported, and no case series has described the safety of this technique in cervical lesions. Herein, we report cases of US-guided CILESIs without fluoroscopy.

**Citation:** Maeda, N.; Maeda, M.; Tanaka, Y. Direct Visualization of Cervical Interlaminar Epidural Injections Using Sonography. *Tomography* **2022**, *8*, 1869–1880. https://doi.org/10.3390/ tomography8040157

Academic Editors: Adam Piorkowski, Rafał Obuchowicz, Andrzej Urbanik and Michał Strzelecki

Received: 21 June 2022 Accepted: 20 July 2022 Published: 22 July 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

21

#### **2. Materials and Methods**

#### *2.1. Patients*

This study was performed at the Maeda Orthopedic Clinic and was approved by its institutional review board (Commission of Ethics approval number 00000003).

Patients provided written informed consent for the cervical epidural block and the publication of this report. Overall, nine patients presented to the Maeda Orthopedic Clinic between August 2021 and January 2022. All patients underwent US-guided CEDI for the treatment of neck pain or interscapular radicular pain due to cervical disc herniation, cervical discopathy, cervical canal stenosis, or thoracic disc herniation.

#### *2.2. US-Guided CEDI*

This novel US-guided CILESIs was performed as follows. First, the axial image of the interspinal ligament, ligamentum flavum, and dura mater as a triple bar sign was identified using a posterolateral with transverse scan approach by a PVI-475BX convex probe (Aplio i800 systems; Canon Medical System, Tochigi, Japan) (Figure 1). To optimize this image, the patient was placed in the prone position, and a pillow was placed under their chest. The patient was then asked to place their forehead on the bed and flex their neck to the maximum extent possible (Figure 2). For needle placement, the tip of the needle was advanced into the ligamentum flavum, tangential to the dura mater, under US guidance using an in-plane technique (Figure 1, Video S1). At first, while inserting and removing the needle, the needle was translocated from the caudal (or cranial) side to the correct plane where the triple bar sign is seen clearly (Figure 3A; view from the side of the body). The needle direction should be corrected if needed. Red arrows indicate an inappropriate needle direction, with an extension line (indicated by dashed arrows) facing the spinal cord or the interspinal ligament. The needle trajectory should be corrected from those indicated by the red arrows to that indicated by the blue arrow (Figure 3B; view from the caudal side of the body). As no critical arteries or nerves are passed using this approach, the needle orbit correction can be repeated just before the needle tip reaches the ligamentum flavum. As a result, accurate needle placement is possible.

**Figure 1.** The illustration shows the needle insertion route on the cross-sectional image of the neck. The main structures are the cord, dura mater, interspinous ligament, and ligamentum flavum. The epidural space exists between the dura mater and the ligamentum flavum. The needle (red arrow) is introduced tangentially in between the ligamentum flavum and dura matter.

**Figure 2.** The patient is placed in the prone position, and the pillow is placed under the patient's chest. Then, the patient is asked to place their forehead on the bed and flex their neck to the maximum extent possible.

**Figure 3.** Orbit correction method.

Video S1. The needle tip is advanced tangentially between the ligamentum flavum and dura matter under ultrasonography (US) guidance. This can be confirmed by the resistance of the needle tip when reaching the complex of the ligamentum flavum felt by the practitioner's hand when checking the US monitor.

While being inserted and removed, the needle is translocated from the caudal (or cranial) side to the correct plane where a triple bar sign is seen clearly (Figure 3A). Next, the needle direction is corrected. Red arrows indicate inappropriate needle directions, with extension lines (indicated by dashed arrows) facing the spinal cord or interspinal ligament. The needle orbit should be corrected from the red to the blue trajectory (Figure 3B, view from the caudal side of the body).

Using a posterolateral approach, the needle is visible from the insertion point to the epidural space between the ligamentum flavum and dura matter. At first, while being inserted and removed (indicated by blue arrows in the figure), the needle is translocated from the caudal (or cranial) side to the correct plane where a triple bar sign (which is composed of the interspinal ligament, ligamentum flavum, and dura matter) is seen clearly (Figure 3A; view from the side of the body). Next, the needle direction is corrected. Red arrows indicate inappropriate needle directions, with extension lines (indicated by dashed arrows) facing the spinal cord or interspinal ligament. The needle orbit should be corrected from the red to the blue trajectory (Figure 3B; view from the caudal side of the body). Because there are no critical arteries and nerves through this approach, the needle orbit correction can be repeated just before the needle tip reaches the ligamentum flavum. As a result, accurate injection is possible.

Confirmation that the needle tip had reached the ligamentum flavum was achieved based on resistance according to the pressure felt by the operator whilst observing the US monitor. We then applied gentle pressure to the syringe plunger as the needle was slowly advanced into the space between the ligamentum flavum and dura mater reaching approximately two-thirds of the triple bar sign, where easy flow of the injectant was facilitated (Figure 4, Video S2). If the needle tip was located more superficially than the ligamentum flavum, any attempt to inject the injectant would be evident in the surrounding muscle tissue or the pseudo-epidural space, such as the space of Okada (Figure 5, Video S3).

**Figure 4.** The illustration shows the cross section of the spinal canal. The main structures are the cord, the dura mater, the interspinous ligament, and the ligamentum flavum. The epidural space exists between the dura mater and the ligamentum flavum. The space of Okada exists between the interspinous ligament and the ligamentum flavum. Ultrasound data show the superb microvascular imaging signal during injection (left picture) and the enlargement of the epidural space after injection (upper illustration and right picture).

Video S2. The needle is gradually advanced while applying pressure to the plunger. If the needle tip is located more superficially than the epidural space, microvascular imaging (SMI) shows that there is no flow of injectant into it. However, once the dense fibrous ligamentum flavum is pierced, the injectant flows. This is immediately confirmed using SMI. Subsequently, the epidural space is gradually enlarged by epidural injection.

Once the dense fibrous ligamentum flavum was pierced, the injectant was able to flow into the epidural space; this was immediately confirmed using superb microvascular imaging (SMI) (Figure 4, Video S2). Lastly, the practitioner verified the epidural space

enlargement using the epidural injection. We used 4 mg of dexamethasone in 6 mL of 0.25% lidocaine for CEDI.

**Figure 5.** The pseudo-epidural space is enlarged by a small dose of injection. The space between the interspinous ligament and ligamentum flavum, such as the space of Okada, is observed sonographically. If the needle tip is located more superficially than the ligamentum flavum, the injectant will appear in the space of Okada or in the spinalis cervicalis muscle once administered.

#### *2.3. CT Epidurography*

To confirm the accuracy of US-guided CEDI, we chose one patient (case 4) and two other patients (a 78-year-old male patient with cervical canal stenosis (Supplementary Materials S1) and a 58-year-old female patient with cervical disc herniation (Supplementary Materials S2) who were not included in this case series because they could not be followed up for >2 months). We performed US-guided CEDI at the C4/5 level. We used a contrast medium instead of dexamethasone (3.5 mL iohexol-240 [Omnipaque-240; GE Healthcare Pharma, Tokyo, Japan]) in 3.5 mL of 0.25% lidocaine. Immediately after injection of the contrast medium, anterior–posterior and lateral epidurograms of the cervical spine were captured. Subsequently, the patients were placed in the prone position within the gantry of the CT scanner (Aquilion Start Canon medical system, Tochigi, Japan) and CT was performed from C1 to L1 5 min after injection. The two orthopedic surgeons and the author have 32, 38, and 24 years of experience, respectively. Technical success was determined by reviewing CT epidurograms for the presence of epidural contrast. The spread of contrast within the epidural space was also assessed on CT epidurograms.

#### *2.4. Pain Intensity*

Pain intensity was evaluated using the visual analog scale (VAS). All patients were instructed on how to use the VAS prior to the cervical epidural block (0, no pain; 10, worst pain conceivable). Before the procedure, an interventional orthopedic doctor questioned the patients regarding their baseline VAS scores for pain. The patients were discharged on the day of the procedure and asked to revisit 1 week and 2 months later for evaluation of side effects and pain scores, respectively. If the pain persisted, we treated patients with an additional block and rechecked their pain scores.

#### *2.5. Functional Ability*

Functional ability was evaluated using the neck disability index (NDI). The NDI is a 10-item self-administered disease-specific questionnaire that evaluates the impact of neck pain on a patient's daily life and the corresponding disability. The questionnaire scores range from 0 to 50; the higher the score, the greater the disability. The validity and reliability of this scale have been previously established [12].

Symptom improvement was assessed before and 2 months after the injection using the VAS and the Japanese version of the NDI. Any potential complications, including headache, increased neck pain, or stiffness, were monitored. Moreover, functional ability was rechecked in patients treated with an additional block.

Patients were asked about their opioid consumption, any additional cervical spine injections, and progression to surgery during the follow up period, and the answers were documented in their profile.

#### **3. Results**

During the study period, nine participants with radicular pain underwent US-guided CEDI. The radicular pain in one, two, three, and three patients was caused by thoracic disc herniation, cervical disc herniation, cervical discopathy, and cervical canal stenosis, respectively. These patients were followed up from 60 to 348 days after the procedure. Their baseline data are presented in Table 1. The baseline and demographic characteristics of the patients were recorded, and the mean age of patients was 51.2 years (range, 23–74 years). Six participants were men and three were women. The mean body mass index was 24.2 km/m<sup>2</sup> (range, 19.3–31.2 km/m2). At the 2-month follow-up, there were significant reductions in NDI from 44.4 to 15.6 points in Table 2. The mean VAS score of neck pain was 8.3 at before the procedure and 1.8 at the 2-month follow-up in Table 2. Among those who reported having recurrent pain (four patients), no patient used opioids for analgesia, four patients reported receiving additional injections, and no patient underwent surgery. During the follow-up period, one patient was able to discontinue opioids.


**Table 1.** Clinical and demographic characteristics.

**Table 2.** Pain intensity and functional ability before and after ultrasound-guided epidural injection.


Vascular or subdural injections were not confirmed in any of the patients during the procedure. Furthermore, we did not encounter any significant complications, including stroke or persistent neurological deficits, during the procedure or 2-month follow-up. Nausea and general fatigue were observed after epidurography using a contrast media (case 4); however, the symptoms resolved within a day.

CT epidurograms revealed that the procedure was technically successful in all three patients from whom they were obtained, including case 4. The spread of contrast was confirmed from the C2 vertebra level to the Th5 vertebra level according to the axial CT epidurogram.

SMI was performed with a small injection. An appropriate epidural space was detected using SMI after the needle was advanced. The epidural space was then enlarged by means of an additional injection directed to the area where the SMI signal was detected (Figure 4, Video S3).

Video S3. The retrodural space of Okada is gradually enlarged by epidural injection. This lesion is not a true space in the epidural space. The loss-of-resistance method cannot differentiate between the space of Okada and the epidural space.

#### *3.1. Representative Cases*

#### 3.1.1. Case 4

A 73-year-old woman presented with severe posterior cervical pain and right shoulder pain. The pain did not resolve with the use of medications (an opioid, an anti-inflammatory analgesic, and a muscle relaxant). Magnetic resonance imaging revealed cervical canal stenosis. Her pain persisted for 3 weeks; therefore, US-guided C3/4 interlaminar CEDI of 1 mL dexamethasone (4 mg) and 6 mL 0.25% lidocaine was administered. In the examination room, SMI was used to confirm that the parenteral solution had spread throughout the epidural space. Subsequently, the VAS score decreased from 10 to 5. An additional four blocks were performed for 2 months, and the VAS decreased from 5 to 2.

After 2 months, she experienced exacerbations and remission of symptoms. She received US-guided epidural blocks 20 times in total. We confirmed the 20th US-guided injection using CT epidurography (Figure 6).

**Figure 6.** CT epidurogram after C4/5 US-guided epidural injection administered to a 73-year-old female with severe posterior cervical pain and right shoulder pain (case 4) associated with cervical canal stenosis. (**A**) Sagittal view of CT epidurogram. Dorsal spread (white arrow heads) between C2 and Th3 are confirmed. (**B**,**C**) Axial view of CT epidurogram (a slice of C4 vertebra and C4/5 disc; injection site). Dorsal spread (white arrow heads) and C5 root filling (white arrows) were confirmed. (**D**,**E**) Axial view of CT epidurogram at C2 and Th3 vertebra ). Cephalad spread (black arrows) was confirmed up to C2 vertebra level and caudal spread (black arrow heads) was confirmed down to Th3 vertebra level.

#### 3.1.2. Case 5

A 73-year-old man presented with severe pain in his right shoulder and arm. Radiographs revealed that the patient had severe cervical spondylosis. Considering that severe night pain and insomnia had persisted for 2 weeks, US-guided C5/6 interlaminar CEDI of 1 mL dexamethasone (4 mg) and 6 mL 0.25% lidocaine was administered. Subsequently, the VAS score decreased from 10 to 1. No additional blocks were administered. No recurrence of pain was observed at the 8-month follow-up.

#### 3.1.3. Case 9

A 23-year-old man presented with severe pain in his right neck and interscapular region. The pain persisted despite diagnostic facet joint blocks (right Th5/6, right C3/4, and right C4/5 facet blocks). Magnetic resonance imaging revealed no vertebral bone injury; however, damage to the C3/4 and C4/5 discs was observed. The pain persisted for 3 months; therefore, US-guided C5/6 interlaminar CEDI of 1 mL dexamethasone (4 mg) and 6 mL 0.25% lidocaine was administered. In the examination room, SMI was used to confirm that the parenteral solution had spread throughout the epidural space. Subsequently, the VAS score decreased from 10 to 1. Two additional blocks were administered over 7 weeks, leading to the resolution of his pain. No recurrence of pain was observed at the 3-month follow-up.

#### **4. Discussion**

Reports of serious complications such as spinal cord injury have led to concerns regarding the safety of CEDI [1–3]. Epidural injections can be performed with or without a guide (blinded injection). Fluoroscopy or CT guidance aims to reduce the risks of epidural injection, such as dural puncture or spinal cord injury [13–16].

In the interlaminar approach, a guide is used to advance the needle tip just before the epidural space [13].

CT guidance in interlaminar ESI improves the evaluation of the proper positioning of the needle owing to its high resolution and cross-sectional imaging [13]. However, the introduction of the needle into the epidural space cannot be visualized according to the movement of the needle during the procedure. US, as an alternative method for performing CILESI, could help reduce the risk of injury by allowing direct visualization of the location of the needle tip during the procedure with cross-sectional imaging. However, a technique for US-guided CILESI has not been reported, and no case series have described its safety because US has some limitations in neuraxial (epidural or intrathecal) procedures as it has a limited resolution at deep levels and near bony surfaces that affect image quality, and it is not possible to visualize the real-time propagation of the injectable in the epidural or intrathecal space [17].

CEDI is based on the LOR method [2,7]. The epidural space is closed by negative pressure unless there is an abscess or bleeding. It is approximately 1–2 mm in diameter at levels above C6-C7 and is less dependent on individual variation, including the presence of spinal stenosis [14,15]. Thus, the safety zone for interlaminar CEDI is very narrow [16]. In reviews of malpractice claims between 2005 and 2008, 64 cases involved cervical interventions, 20 of which resulted in direct spinal cord injury associated with interlaminar CEDI [5,9].

The needle tip position must be visualized throughout the procedure to ensure accurate injection into the epidural space. Clinical practice guidelines recommend fluoroscopic guidance in all CEDIs [18]. However, fluoroscopic guidance methods are limited because they use bones as landmarks. Therefore, it is difficult to guide the needle tip into the gap of the ligamentum flavum and dura mater safely, as it is approximately 1 mm thick and permeable to X-rays, without direct visualization.

For the accurate insertion of the needle, the US monitor, the affected area, and the practitioner should be in a straight line. However, this is not always possible during outpatient examinations because the size of the examination room is limited. Furthermore, since the needle is inserted deeply, needle orbit correction is often needed. As there are no

critical arteries or nerves beyond the ligamentum flavum through this approach, the needle orbit correction can be repeated just before the needle tip reaches the ligamentum flavum. As a result, accurate injection is possible with our US guided method.

The LOR technique relies on penetration of the ligamentum flavum. However, up to 74% of the ligamenta flava are discontinuous in the midline of the cervical region [19]. Dural puncture cannot be prevented based on bony landmarks or the LOR technique because the needle is inserted near the midline under conventional method. Furthermore, the safety of the CEDI procedure is highly dependent on the operator's experience. In contrast, US can easily identify the ligamentum flavum, dura mater, and the needle tip independent of the operator's experience.

If the needle is advanced deeply, dural puncture or spinal injury can be avoided because the needle is tangentially advanced with respect to the dura. Even in cases of inappropriate injection, the practitioner could use the US image to confirm and terminate the inappropriate injection from the beginning, if necessary. Therefore, serious complications, such as total spine anesthesia or spinal injury, could be avoided. The direction of the needle has two advantages. The first is improved needle visibility by bringing the needle closer to the probe in parallel. The second is the avoidance of dural puncture and spinal cord injury.

Unlike US guidance, fluoroscopy and CT guidance (Figure 7) cannot accurately visualize the needle tip position. Fluoroscopy-guided CEDI can only confirm whether the LOR method is performed correctly based on contrast, identifying the epidural space after the trial of LOR. Additionally, CT-guided CEDI cannot show needle movement in real-time as the procedure progresses. The placement of the needle tip is only confirmed after needle insertion. SMI methods have been used to identify slow blood flow in the body. Unlike CT and fluoroscopy, SMI helps visualize not only microvessels with slow blood flow but also the movement of liquid (injectant) flow without contrast injection [20], and it can be a useful tool to confirm if the needle tip is in proper anatomical position e.g., the epidural space. Under SMI, the spread of the injectant can be visualized using US while simultaneously confirming the proper placement of the tip of the needle between the ligamentum flavum and dura matter (Figure 3) [6]. The major difference is that it offers real-time, radiation-free guidance for interventions.

The LOR method can also be used to precisely detect the epidural space using SMI signals. However, the LOR method is associated with false-positive LOR before the needle enters the epidural space. False-positive LOR occurs in 29.4% of patients who undergo CEDI under conventional fluoroscopic guidance [7]. The retrodural space of Okada is a potential space dorsal to the ligamentum flavum that allows communication between the bilateral facet joints and interspinous bursa [8,21,22]. In the current study, the space of Okada on the ligamentum flavum was visible using US (Figure 5, Video S3).

US guidance allows better positional relationships between the needle tip and anatomical structures (e.g., lamina, space of Okada, and epidural space) in cross-sectional imaging such as CT guidance. A previous investigation examining conventional fluoroscopy-guided lumbar interlaminar ESI demonstrated that many non-target injections in the retrodural space of Okada are likely to go unnoticed at the time of the procedure [21]. This may be because conventional fluoroscopy contrast material in the retrodural space of Okada can mimic true dorsal epidural spread as these two spaces run parallel to each other and may therefore overlap in standard lateral and anteroposterior projections [21,22]. Our US data also revealed that the space is too close to be identified by fluoroscopy. This suggested that inappropriate injection by LOR could be avoided using US.

Riveros-Perez et al. introduced color Doppler imaging to confirm the correct position of the epidural needle in the lumbar spine [23]. However, this method has not been applied to CEDI [17]. For cervical epidural injection, Zhang et al. reported transforaminal epidural steroid injections using sonography [24]. However, they did not employ color Doppler imaging to confirm the correct position of the needle and neither did they visualize the spinal cord, its nutrient vessels, or the radicular artery during the procedure. Serious complications such as spinal cord injury, either directly or indirectly via injury of its nutrient vessels, have been reported in transforaminal epidural steroid injections [1–3]. We visualized the spinal cord and surrounding vessels during the interlaminar epidural injection using the SMI method. To the best of our knowledge, our study is the first to achieve direct visualization of not only the entry route of the needle but also of the spinal cord and surrounding vessels during the CILESIs procedure.

**Figure 7.** Upper figures show our cervical epidural injection (CEDI) method in which the needle is inserted tangentially into the dura (**A**–**E**). Lower figures show the conventional CEDI method wherein the needle is inserted nearly perpendicular to the dura (**F**–**J**). In both methods, loss of resistance may be confirmed before and after the insertion into the flavum (**A**,**C**,**F**,**H**). The closer the needle is inserted vertically, the higher the risk of dural puncture because the safety margin between the flavum and the dura becomes narrower (**G**,**H**). The risk of dural puncture can be avoided if the needle is inserted tangentially into the dura (**C**,**D**). In the conventional method (**J**), when the needle is advanced deeply, dural puncture cannot be avoided. Meanwhile, in our method, dural puncture can be avoided because the needle penetrates tangentially to the dura (**E**). The flow of the injectant into the epidural space is visualized in real-time under ultrasound-guided epidural injection (**C**).

#### **5. Conclusions**

In conclusion, we propose the possible use of US for easier and safer CILESIs compared with conventional methods. US for the direct visualization of CILESIs may allow for safer outpatient injections.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/tomography8040157/s1, Figure S1: a 78-year-old male patient with cervical canal stenosis, Figure S2: a 58-year-old female patient with cervical disc herniation, Video S1: Approach of US guided cervical epidural injection, Video S2: Enlargement of the space of OKADA, Video S3: Spread of the epidural space during epidural injection.

**Author Contributions:** Conceptualization, M.M. and N.M.; methodology, N.M.; investigation, M.M.; resources, M.M.; data curation, M.M.; writing—original draft preparation, N.M.; writing—review and editing, Y.T.; visualization, M.M.; supervision, Y.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** This study was performed at the Maeda Orthopedic Clinic and was approved by its institutional review board (Commission of Ethics), 00000003.

**Informed Consent Statement:** The patients provided written informed consent for a cervical epidural block and the publication of this report.

**Data Availability Statement:** The data that supports the findings of this study are available in the supplementary material of this article.

**Acknowledgments:** The authors appreciate Ichiro Higashiyama for correcting data of CT-epidurography.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Article* **Measures of Corticalization**

**Marcin Kozakiewicz**

Department of Maxillofacial Surgery, Medical University of Lodz, 113 Zeromskiego Str., 90-549 Lodz, Poland; ˙ marcin.kozakiewicz@umed.lodz.pl; Tel.: +48-42-6393422

**Abstract:** After the insertion of dental implants into living bone, the condition of the peri-implant bone changes with time. Implant-loading phenomena can induce bone remodeling in the form of the corticalization of the trabecular bone. The aim of this study was to see how bone index (BI) values behave in areas of bone loss (radiographically translucent non-trabecular areas) and to propose other indices specifically dedicated to detecting corticalization in living bone. Eight measures of corticalization in clinical standardized intraoral radiographs were studied: mean optical density, entropy, differential entropy, long-run emphasis moment, BI, corticalization index ver. 1 and ver. 2 (CI v.1, CI v.2) and corticalization factor (CF). The analysis was conducted on 40 cortical bone image samples, 40 cancellous bone samples and 40 soft tissue samples. It was found that each measure distinguishes corticalization significantly (*p* < 0.001), but only CI v.1 and CI v.2 do so selectively. CF or the inverse of BI can serve as a measure of peri-implant bone corticalization. However, better measures are CIs as they are dedicated to detecting this phenomenon and allowing clear clinical deduction.

**Keywords:** dental implants; long-term results; long-term success; functional loading; peri-implant bone; intra-oral radiographs; radiomics; texture analysis; corticalization; bone remodeling

#### **1. Introduction**

The functional loading of dental implants induces permanent changes in the alveolar crest [1,2]. The functional loading of intraosseous dental implants causes significant changes in the structure of the alveolar marginal bone, observed radiographically [3]. There was corticalization and associated marginal bone loss relentlessly progressing over the five and ten years of observation presented previously [3]. It is expressed in the loss of trabeculation (lower entropy of bone radiostructure) in favor of the unification of the arrangement of bone components and their massification (increase of long elements in radiograph). Both of these structural changes are summarized in the bone index (BI). The conducted analysis strongly suggests that the phenomenon of corticalization is a nonbeneficial alteration of the bone around the implants (at least in the scope disclosed in this study). It means that marginal bone loss will increase as corticalization progresses.

The trabecular structure disappears and is successively replaced by cortical bone-like tissue. These observations were made on the digital analysis of peri-implant bone structure in intraoral radiographs. For this, the bone index [4,5] was used, or, strictly speaking, the inverse of this index since BI is used to detect trabecular bone. Due to the fact of dichotomous deduction possibilities (cancellous bone vs. cortical bone), 1/BI was proposed for detecting corticalization.

However, BI is oriented toward detecting cancellous bone. In trabecular structure radiographs, BI reaches the highest values. In contrast, it reaches low values in other bone structures. The author suspects that low BIs occur not only in images of cortical and corticalized bone but also in areas of bone atrophy (uniformly radiologically translucent). This suspicion is related to the structure of the BI [4,6] since there is a measure in its structure that highlights the existence of long strings of pixels of similar brightness (in other words, of similar radiographic translucency). This measure is not related to high brightness

**Citation:** Kozakiewicz, M. Measures of Corticalization. *J. Clin. Med.* **2022**, *11*, 5463. https://doi.org/10.3390/ jcm11185463

Academic Editors: Michał Strzelecki, Rafał Obuchowicz, Andrzej Urbanik and Adam Piorkowski

Received: 9 July 2022 Accepted: 15 September 2022 Published: 16 September 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

(optical density) only but shows both high and low optical density regions. BI cannot be the only measure for evaluating bone at dental implant site. Unfortunately, BI does not indicate whether a clinically suspicious site is corticalization or bone loss. This is a significant disadvantage. To avoid it, each examined site or radiograph should be subjected to visual inspection, which precludes the automation of the analysis and completely excludes the use of radiomics.

The aim of this study was to see how BI behaves in areas of bone atrophy (radiologically translucent non-trabecular areas) and to propose other indices specifically dedicated to detecting corticalization in living bone.

#### **2. Materials and Methods**

The source of the scientific material included in this study was digital intraoral radiographs [7] taken with the Digora Optima system (Soredex, Helsinki, Finland): 7 mA, 70 kV an 0.1 s (Focus apparatus—Instrumentarium Dental, Tuusula, Finland). Positioner Rinn (Densply, Charlotte, NC, USA) was used for the 90◦ angle of X beam to the surface of phosphor plate. Storage phosphor plates were read immediately after exposure.

Square areas of 3844 pixels (62 × 62), i.e., regions of interest (ROIs) in 8-bit, greyscale images were included in the study, numbering 40 for the compact (cortical) bone images, 40 for the cancellous (trabecular) bone images and 40 soft tissue images (Figure 1). A total of 120 ROIs were analyzed.

**Figure 1.** Regions of interest were located in cortical bone (ROI 1), trabecular bone (ROI 2) and soft tissue (ROI 3) in main window of MaZda. Next, a series of textural features was extracted (MZ Reports—on the left) and exported in comma-separated vector format (CSV).

This provided information on three unique regions: cortical bone, trabecular bone and soft tissue. The textures of the X-ray images were analyzed in MaZda 4.6 freeware invented by the University of Technology in Lodz [8] to test measures of corticalization in control environments of trabecular bone (representing original bone before implantdependent alterations) and soft tissue (representing product of marginal bone loss). MaZda provides both first-order (mean optical density) and second-order (entropy, differential entropy (DifEntr), long-run emphasis moment (LngREmph)) data. Due to the fact that the second-order data are given for four directions in the image, and in the present study, the author did not wish to search for directional features, the arithmetic mean of these four primary data was included for further analysis. The regions of interest (ROIs) were normalized (*μ* ± 3*σ*) to share the same average (*μ*) and standard deviation (*σ*) of optical density within the ROI. To eliminate noise, [9] worked on data reduced to 6 bits. For analysis in a co-occurrence matrix, a spacing of 5 pixels was chosen. In the formulas that follow, *p*(*i*) is a normalized histogram vector (i.e., histogram whose entries are divided by the total number of pixels in ROI), *i* = 1,2,..., *Ng* denotes the number of optical density levels. Mean optical density (only a first-order feature) is calculated as follows:

$$
\mu = \sum\_{i=1}^{N\_3} i p(i) \tag{1}
$$

Second-order features are found by:

$$Entropy = -\sum\_{i=1}^{N\_\mathcal{S}} \sum\_{j=1}^{N\_\mathcal{S}} p(i, j) \log(p(i, j)) \tag{2}$$

$$DiffEntr = -\sum\_{i=1}^{Ng} p\_{x-y}(i) \log \left( p\_{x-y}(i) \right) \tag{3}$$

where Σ is sum, *Ng* is the number of levels of optical density in the radiograph, *i* and *j* are optical density of pixels 5-pixel distant one from another, *p* is probability and *log* is common logarithm [10]. The differential entropy calculated in this way is a measure of the overall scatter of bone structure elements in a radiograph. Its high values are typical for cancellous bone [4,11,12]. Next, the last primary texture feature was calculated (Figure 2):

$$LngREmph = \frac{\sum\_{i=1}^{N\_\mathcal{S}} \sum\_{k=1}^{Nr} k^2 p(i,k)}{\sum\_{i=1}^{N\_\mathcal{S}} \sum\_{k=1}^{Nr} p(i,k)} \tag{4}$$

where Σ is sum; *Nr* is the number of series of pixels with density level *i* and length *k*; *Ng* is the number of levels for image optical density; *Nr* is the number of pixel in the series; and *p* is probability [13,14]. This texture feature describes thick, uniformly dense, radio-opaque bone structures in intraoral radiograph images [4,12].

The equations for DifEntr and LngREmph were subsequently used for the index construction [4–6,12]. The bone index (BI), which represents the ratio of the diversity of the structure observed in the radiograph to the measure of the presence of uniform longitudinal structures, was calculated:

$$Bone\ Index = \frac{DiffEntr}{LngREmph} \tag{5}$$

Two more formulas were developed with the intention that they would describe the intuitive increases in their values together with the progression of corticalization and that they would suppress the results for cancellous and soft-tissue sites by representing such sites with low values:

$$\text{Corticalization Index } var.1 = \frac{\text{LongREmph} \cdot \text{Mean}}{DiffEnttr} \tag{6}$$

$$\text{Criticalization Index } ver.2 = \frac{\text{Lng} \,\text{REmph} \cdot \text{Mean}}{\text{Entropy}} \tag{7}$$

The Kruskal–Wallis test was used for the comparison of medians between cortical, and trabecular or soft tissue radiograph (Statgraphics–StatPoint Technologies, Inc., The Plains, VA, USA). Factor analysis was used to find the statistically supported next measure for the corticalization process product. Input vectors: mean optical density, texture entropy, DifEntrp, and LngREmph. The procedure was performed for factors of eigenvalue ≥ 1. A probabilistic neural network (PNN) to classify cases into different ROI was applied. Rate of correctly classified ROIs by the network was evaluated.

**Figure 2.** The source material and the primary texture features extracted from it. The meanings of the ROIs are the same as in Figure 1. Maps of the local intensity of the studied features are below the original radiographs. The map is created from square boxes of nine pixels. In the maps of features, lighter areas indicate higher local intensity of the feature, while darker areas indicate lower intensity of the feature.

#### **3. Results**

Calculations for selected measures of corticalization in radiographs of three types of tissue representing the corticalization phenomenon, bone loss and the reference region of cancellous bone are shown in Table 1. The results of the primary bone imaging features are shown in Figure 3. These consisted of one first-order feature (mean optical density) and two second-order features (DifEntr and LngREmph).

**Table 1.** Numerical results of the investigation of selected measures of corticalization.


<sup>1</sup> Statistically significant difference found between all the ROIs compared with each other; <sup>2</sup> ROI 3 is significantly lower than ROI 1 as well ROI 2; <sup>3</sup> ROI 3 is significantly higher than ROI 1 as well as ROI 2; <sup>4</sup> ROI 1 is significantly higher than ROI 2 as well as ROI 3. ROI—region of interest; LngREmph—long-run emphasis moment.

**Figure 3.** Based on the above three features (DifEntrp, Mean optical density, LngREmph), the algorithm manages to initially separate the results for the three tissues (ROIs), but corticalization (Cortical) is not well discriminated here. It is worth noting that the simple measure of mean optical density itself shows the differences between the regions of interest studied.

The constructed indices were then examined in three ROIs: bone index (Figure 4), corticalization index ver. 1 (Figure 5) and corticalization index ver. 2 (Figure 6).

**Figure 4.** The bone index (BI) was calculated for the detection of normal bone (i.e., trabecular bone) within dental alveolus during guide bone regeneration. That is why BI reaches the highest values in ROI 2 representing trabecular bone. There are significant statistical differences between each ROI.

**Figure 5.** Corticalization index ver. 1 (CI v. 1) is based on two components included in BI and mean optical density. The components are arranged inversely to the BI to emphasize the corticalization sites rather than trabeculation, and the mean optical density enhances this effect because it is located in the numerator and is highest in the cortical bone.

**Figure 6.** Corticalization index ver. 2 (CI v. 2). This corticalization measure differs from version 1 by replacing differential entropy (ver. 1) in the denominator with entropy (here, ver. 2). This was dictated by the good statistical separation of ROI 1 from the other two ROIs by entropy. However, due to the greater spread of entropy in ROIs than differential entropy, the separation between ROIs is weaker here (but still highly statistically significant: *p* < 0.001).

The purpose of the factor analysis was to obtain a small number of factors that would account for most of the variability in the four textural features (mean optical density, texture entropy, DifEntrp, and LngREmph). In this case, the factor was extracted with a high eigenvalue, 3.30 (much greater than or equal to 1.0). It accounted for 82.4% of the variability in the original texture data. Since principal components was selected, the initial communality estimates were set to assume that all of the variability in the data was due to common factors. Moreover, the Kaiser–Meyer–Olkin measure of sampling adequacy (KMO) was above 0.6 for that set of input features. This factorability test indicates whether or not it is likely to be worthwhile attempting to extract factors from a set of variables. The KMO statistic provides an indication of how much common variance is present. For factorization to be worthwhile, KMO should normally be at least 0.6. Since KMO = 0.768, factorization was likely to provide interesting information about any underlying factors. The equation that estimated the common factor (the corticalization factor, CF) was performed to represent the factor loadings:

$$CF = 0.8446 \cdot Mean + 0.9555 \cdot Entropy + 0.9066 \cdot DiffEntr - 0.9211 \cdot Long \, REmph \tag{8}$$

where the values of the variables in the equation are standardized by subtracting their means and dividing by their standard deviations. It also shows the estimated communalities, which can be interpreted as estimating the proportion of the variability in each variable attributable to the extracted factors.

Factor analysis indicated that by placing the main emphasis on the simple measurement of mean optical density and measuring the amount of chaoticity in the texture, it is possible to more than adequately detect corticalization sites in the bone image (high CF = 114 ± 23) with simultaneous indication of normal trabecular structure (intermediate values 80 ± 12) and sites that are no longer bone, such as those affected by marginal bone loss (lower values 29 ± 14). The presence of pixel long series of similar optical density is minimized in this corticalization evaluation technique. However, removing LngREmph from the analysis lowers the KMO to 0.618. Thus, one should suspect that short pixel series (i.e., the inverse of LngREmph) is more important in assessing corticalization. A second conclusion from this relationship is the essentiality of evaluating pixel series for indicating corticalization sites.

Thus, the last of the corticalization measures examined here was obtained from factor analysis: CF (Figure 7). It was strongly stratified and allowed for good discrimination of cortical bone from cancellous bone, cancellous bone from soft tissue, and soft tissue from cortical bone (*p* < 0.001).

**Figure 7.** Corticalization Factor (CF). It has statistical features similar to BI, but it is most strongly expressed in cortical sites, weaker in trabecular bone and weakest in soft tissues.

The relationships of the corticalization index with the bone index and corticalization

**Figure 8.** The relationships of selected corticalization measures in the evaluation of intraoral radiographs. (**a**) Corticalization index ver.1 with bone index (BI). A probabilistic neural network (PNN) used to classify cases into different three structures in radiograph (cortical, trabecular, bone loss), based on two input variables (corticalization index ver.1 and bone index). Of the 120 ROIs, 93% were correctly classified by the network. (**b**) Corticalization index ver.1 with corticalization factor. Among the 120 ROIs used, 94% were correctly classified in this pair of corticalization evaluators.

#### **4. Discussion**

It is worth noting that a simple measure derived from radiograph histogram analysis (i.e., mean optical density) has been used in dentomaxillofacial radiology for decades [9,15–24]. It carries a great deal of clinically useful information but does not allow the automation feature of radiomics [25–28] because it requires an analog context for understanding the significance of local density changes. A second argument is the vast amount of other information coming from the radiograph beyond the histogram data. In dentistry, more than 10,000 texture features computing from the determined ROIs are now possible [29]. A final issue is the non-specificity of the mean optical density for assessing bone corticalization since residual granules of biomaterial previously implanted into bone, for example, can be detected by this feature [30,31].

In peri-implant bone, optical density increases on plain intraoral radiography in patients treated with immediate-loading implants [1,32]. Similar observations were made for late-loaded implants in the same time horizon (12 months of functional loading) that were noted in the peri-implant bone texture structure [33]. The sum of squares, a feature from the co-occurrence matrix, was studied, and it was found that there is a significant decrease in the value of this texture feature around the integrated dental implant at 12 months after prosthetic loading. This indicates a homogenization of the bone texture and a decrease in its intrinsic contrast [33]. These are the initial reports describing the phenomenon of corticalization of the alveolar crest caused by dental implants.

How to measure the product quantity derived from corticalization process in peri-implant bone in a clinical situation is a critical question. In recent years, the occurrence of corticalization in peri-implant bone was mentioned in scientific literature [32,34–36], and attempts are being made to describe this phenomenon [3] and get to know its clinical significance.

The importance of the standardization of images and ROI should be emphasized first. The approach to this issue depends on the tools used later. When digital radiographic subtraction is used [7,9,16–18,21,22], geometric alignment is necessary first because two radiographs of the same implant but taken at different times are superimposed. Rotational, translational, scale and affine distortions need to be corrected. Next, alignment and contrast brightness are needed. This is best achieved by aligning the histograms of the reference locations [20,23]. In the research presented in this study, this second range of alignment is essential. An alignment algorithm is introduced in the MaZda program consists of standardizing the ROI. The ROIs were adjusted (μ ± 3σ) to share the same average (μ) and standard deviation (σ) of optical density within the ROI.

Bone index (BI) is a good measure for determining the qualitative changes occurring in the cancellous bone of the jaw. A decrease in its value indicates the disappearance of the structure characteristic of trabecular bone. This is most likely related to corticalization since low BI precedes the appearance of marginal trabecular bone in dental implantology by years. There was a strong association of low BI (0.41 ± 0.19) present in the fifth year of implant use with marginal bone loss at that time (*p* < 0.0001) [3]. Bone index also well describes the results of guided bone regeneration inside the alveolus [4]. It highlights the appearance of bone trabeculae at the site of biomaterial implantation. Inferences based on the BI also seem to work well in other disciplines, i.e., in the analysis of bone consolidation quality [37], where a low BI is present at post-fracture sites because the islands of bone densities are more homogeneous (compact) than cancellous bone. This results in a less-chaotic structure, i.e., entropy is reduced (and this entropy is the numerator of the fraction forming the BI). For the same reasons, the bone image here has broad and uniform radio-opaque fields, where long lines of pixels of the same optical density can be found. This causes a high LngREmph, and this is the denominator of the fraction that forms BI. Thus, it affects the reduction of the final BI to 0.70–0.79. Unfortunately, LngREmph (as well BI) cannot describe the pixel series as high optical density (bone apposition) or low optical density (bone loss) points in an image (BI is also low in homogenous radiologically translucent regions).

Moreover, in a study of corticalization, one would need a tool indicating the sites of corticalization rather than the inverse, a tool indicating trabeculation, since bone loss is much more strongly represented in this index (i.e., 1/BI) than corticalization itself. Corticalization index ver. 1 is based on up to the inverse of the described bone index

enhanced by the mean optical density result in the ROI. Corticalization index ver. 2 in its design is similar to the inverse of the texture index described previously [5].

The presence of statistically significant differences between the ROIs of the corticalization measures studied indicates that they are all useful to some degree. However, the three tissues tested differed from each other in mean optical density, differential entropy, bone index, corticalization factor). This does not provide a simple measure indicating the site of corticalization searched for in this study. Nevertheless, it should be stated here that both mean optical density and corticalization factor are highest at the site of corticalization. Entropy, on the other hand, is uniformly elevated in both bone tissues (i.e., corticalized and trabecular) relative to soft tissue, which is almost lacking a chaotically scattered texture pattern (Table 1). The next one-element group is the long-run emphasis moment (LngREmph), whose value is highest in soft tissue and lowest in both bone types studied (*p* < 0.001). Yet other measures studied here (both corticalization indices) unambiguously indicate that a given site is corticalized and significantly different from both trabecular bone and well bone loss (i.e., soft tissue). Thus, these two measures do not have the interpretive contamination of random bone loss detection introduced inside. High values here indicate only a corticalization phenomenon, in contrast to the low values, which indicate everything else, i.e., trabecular bone and bone loss. Therefore, these two measures also cannot be used for evaluating the results of guided bone regeneration (the bone index is great for this purpose), nor can they be used to study sites of bone loss. It will be possible in the future to select the best measure for studying a particular phenomenon in peri-implant bone. However, for the considerations in this study, the indices of corticalization (CI ver.1 and CI ver. 2) are the most interesting measures selected.

It seems appropriate to present a dedicated index for detecting corticalization as a phenomenon important for the long-term success of implant treatment. The purpose of this paper is to present such an index for clinical application. The interpretation inconvenience of 1/BI can be resolved by associating the index with high brightness (i.e., high optical density) as typically found in radiographic images of compact (cortical) bone. This gave the idea to include in the index the first-order feature of mean brightness/optical density, i.e., the mean from the histogram of the examined region of interest.

Both versions of CI allow for distinguishing corticalization from atrophy and, of course, also from trabecular bone. The question is whether it is better to rely on BI in the assessment of corticalization, which after all is designed to search for trabecular bone in X-ray images and which also distinguishes bone atrophy well (BI for cancellous bone is the highest, for cortical bone, it is statistically significantly lower and for the site of bone loss, it is significantly lower to corticalized bone). Or is it better to rely on a CI that indicates only corticalization and describes trabecular bone and bone loss together at an equal level i.e., CI ver.1 approx. 114, CI ver.2 approx. 52 for non-corticalization sites.

When considering these aspects of the analysis, it should be emphasized that marginal bone loss has already been very well described in the literature, and the methods to diagnose it are known and evident [38–43]. It is also known to be an unequivocally unfavorable phenomenon for implant success [44–46]. Corticalization itself is suspected as a potentially unfavorable prognostic factor [3]. Thus, it seems best at this stage of the understanding of this phenomenon to focus on the dichotomous separation of the peri-implant bone image structure: corticalized versus other (i.e., cancellous or affected by loss).

As is well known, cortical structural changes can have very serious adverse effects, as in patients treated with antiresorptive therapy (MRONJ) [47], less severe as it seems after a decade of monitoring bone transformation around dental implants [3] or perhaps have a positive effect as in the case of immediate loading of dental implants [32].

The phenomenon described here is so pronounced in the jaws because the bone appositional index here is one of the highest in the body [48]. It is certainly higher than in the iliac bone, femur or vertebrae. In the mandible, the bone apposition rate is 0.086–0.088 μm per hour. This process guarantees the osseointegration of the dental implants in the first phase but is probably responsible in later years for the corticalization of the surrounding area modulated by the permanent loading of the bone by biting and chewing forces. The remodeling and the superimposition of new osteons to the older [49]. This late effect at sufficiently high levels can lead to increased bone fragility and brittleness through the mechanism of osteon hypermineralization, which is related to the process of bone apposition [50]. However, in the jawbone, the phenomenon occurs on a microscale. Whether it is a negative process for the long-term maintenance of functioning dental implants is uncertain. Perhaps for this reason, it is worth thinking about other methods of building measures to describe bone corticalization.

A new approach is the use of factor analysis to evaluate peri-implant bone. The hopes placed in it are based on good predictive experiences from other surgical teams [51,52]. The advantage is the statistically reliable combination of information flowing from several texture features into a single number (corticalization factor) describing the variation occurring in, for example, four features: mean optical density, the frequency of long series of pixels of similar optical density (LngREmph) and two measures of texture pattern scattering (entropy and differential entropy). This aforementioned reliability relies on high eigenvalues and KMO statistics.

Information describing the triplet variation (cortical vs. trabecular vs. bone loss) can reduce the number of features while retaining their internal information by the factor analysis and equation for the corticalization factor (8). It turns out that based on the factor calculated in this way (CF), statistically significant differences (*p* < 0.05) can be indicated between corticalization site roentgenograph (value approx. 120) and cancellous bone (value about 80) versus bone loss (value less than 40). Both analyses (Figure 8) indicate the possibility of the direct transition of cancellous bone to atrophy and of cortical bone to atrophy. Based on the presented methods for the detection of corticalization, there is no indication of a transition state between cancellous bone and bone loss. It is certainly not a corticalized bone. Bone loss can arise directly from one or the other bone tissue.

On the other hand, when considering the interpretive convenience in a study of only corticalization, it is more comfortable to use a tool that gives dichotomously differentiated results, i.e., yes or no for corticalization, and such measures are presented above: the corticalization indices.

It is important and interesting to validate the corticalization measures presented here on a wide range of patients in variable clinical situations (e.g., different implants [53,54], different surgical protocols [55], bone compression screws [56,57], different prosthetic work [58], vitamin D3 levels [47] and molecular signaling modulation [59]). This will allow for choosing the best applications for particular measures. Or perhaps it will prove advisable to use several measures simultaneously, e.g., for monitoring remodeling of the cancellous and cortical substance. It is also important to test the usefulness in other parts of skeletal surgery such as hand [60], foot [61], thoracic [62], orthognathic [63], spine [64], joint replacements [35,65] and rehabilitation [66,67].

One should not forget other measures of corticalization being developed in dentistry itself like fractal dimension and multifractal spectra [53,68–70]. It seems that this valuable and interesting source may in the future bring very useful measures of peri-implant bone remodeling.

This study provides important clinical considerations for dentistry (especially dental implantology). First, it systematizes the possibilities of assessing bone remodeling. It will be possible in the future to select the most suitable index in relation to the observed bone remodeling processes. Second, it is important to relate an objective measure of bone condition with the prediction of dental implant maintenance in normal function. The proposed measures of corticalization are applicable for monitoring bone health around the dental abutments associated with bridges and crowns and the results of guided bone regeneration and tissue bone regeneration in implantology and perioodontology. The clinical relevance of this study can also be seen in monitoring antiresorptive therapy in the treatment of osteoporosis and controlling the metastasis of malignant tumors to bone. Orthopedists, neurosurgeons and hand surgeons who also use metal stabilizers, screws, cages and joint replacements may also benefit. Many fields of medicine need evaluators that assess the condition of bone and its transformation as a function of time. In these example fields of medicine, for example, the use of the corticalization index is being looked at.

#### **5. Conclusions**

The corticalization factor, or inverse of bone index, can serve as a measure of periimplant bone corticalization. However, better measures are the corticalization indices as these are dedicated to detecting corticalization and allowing for clear clinical deduction.

**Funding:** This research was funded by the Medical University of Lodz (grant numbers 503/5-061- 02/503-51-001-18, 503/5-061-02/503-51-001-17 and 503/5-061-02/503-51-002-18).

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Ethics Committee of the Medical University of Lodz, PL (protocol no. RNN 485/11/KB and date of approval: 14.6.2011).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data on which this study is based will be made available upon request at https://www.researchgate.net/profile/Marcin-Kozakiewicz access on 30 June 2022.

**Acknowledgments:** The author would like to thank Andrzej Materka and Michał Strzelecki (Lodz University of Technology, Poland) very much for his valuable comments and consultation of image analysis techniques.

**Conflicts of Interest:** The author declares no conflict of interest.

#### **References**

