Kappa Values For Interobserver Variability Research Articles

Validation of the modified AO sternum classification system.

The Arbeitsgemeinschaft für Osteosynthesefragen (AO) foundation along with the Orthopaedic. Trauma Association (OTA) introduced a new classification for sternal fractures in 2018 aiming to provide greater uniformity and clinical utility for the surgical community. A previous validation study identified some critical issues such as the differentiation between type A and B fractures and localization of the fracture either in the manubrium or in the body. Due to the moderate agreement in inter- and intra-observer variability, some modifications were proposed in order to improve the performance of the classification. The aim of this study was to re-assess the inter- and intra-observer variability after adding modifications to the classification. Our hypothesis was that a significative improvement of inter- and intra-observer variability could be achieved. Twenty computed tomography (CT) scans of patients with sternal fractures were analyzed by six. Junior and six senior surgeons independently. Two assessments were performed with an interval of 6weeks. The kappa (K) value was calculated in order to assess inter- and intra-observer variability. The overall mean kappa value for inter-observer variability improved from 0.364 to 0.468 (p < 0.001). Inter-observer variability mean for location was 0.573 (SD 0.221) and for type was 0.441 (SD: 0.181). Intra-observer variability showed a mean of 0.703 (SD: 0.153) with a statistic significant improvement when compared to the previous study (mean 0.414, SD: 0.256, p < 0.001). By modifying the AO/OTA classification of sternal fractures, the inter- and intra-observer variability improved and now shows moderate to substantial agreement.

Mirels' score for upper limb metastatic lesions: do we need a different cutoff for recommending prophylactic fixation?

HypothesisThe aim of this study was to investigate the reproducibility, reliability, and accuracy of Mirels’ score in upper limb bony metastatic disease and validate its use in predicting pathologic fractures.MethodsForty-five patients with upper limb bony metastases met the inclusion criteria (62% male 28/45). The mean age was 69 years (SD 9.5), and the most common primaries were lung (29%, 13/45), followed by prostate and hematological (each 20%, 9/45). The most commonly affected bone was the humerus (76%, 35/45), followed by the ulna (6.5%, 3/45). Mirels’ score was calculated in 32 patients; with plain radiographs at index presentation scored using Mirels’ system by 6 raters. The radiological aspects (lesion size and appearance) were scored twice by each rater (2 weeks apart). Intraobserver and interobserver reliability were calculated using Fleiss’ kappa test. Bland-Altman plots compared the variances of both individual components and the total Mirels’ score.ResultsThe overall fracture rate of upper limb metastatic lesions was 76% (35/46) with a mean follow-up of 3.6 years (range 11 months-6.8 years). Where time from diagnosis to fracture was known (n = 20), fractures occurred at a median 19 days (interquartile range 60-10), and 80% (16/20) occurred within 3 months of diagnosis.Mirels’ score of ≥9 did not accurately predict lesions that fractured (fracture rate 11%, 5/46, for Mirels’ ≥ 9 vs. 65%, 30/46, for Mirels’ ≤ 8, P < .001). Sensitivity was 14%, and specificity was 73%. When Mirels’ cutoff was lowered to ≥7, patients were more likely to fracture than not (48%, 22/46, vs. 28%, 13/46, P = .045); sensitivity rose to 63%, but specificity fell to 55%.Kappa values for interobserver variability were κ = 0.358 (fair, 95% confidence interval [CI] 0.288-0.429) for lesion size, κ = 0.107 (poor, 95% CI 0.02-0.193) for radiological appearance, and κ = 0.274 (fair, 95% CI 0.229-0.318) for total Mirels’ score. Values for intraobserver variability were κ = 0.716 (good, 95% CI 0.432-0.999) for lesion size, κ = 0.427 (moderate, 95% CI 0.195-0.768) for radiological appearance, and κ = 0.580 (moderate, 95% CI 0.395-0.765) for total Mirels’ score.ConclusionsThis study demonstrates moderate to substantial agreement between and within raters using Mirels’ score on upper limb radiographs. However, Mirels’ score had a poor sensitivity and specificity in predicting upper extremity fractures. Until a more valid scoring system has been developed, based on our study, we recommend a Mirels’ threshold of ≥7/12 for considering prophylactic fixation of impending upper limb pathologic fractures. This contrasts with the current ≥9/12 cutoff, which is recommended for lower limb pathologic fractures.

Interobserver variability in high-resolution CT of the lungs

PurposeTo quantify the interobserver variability among the most frequently encountered parenchymal patterns in High Resolution CT (HRCT) and to compare the interobserver variability in the application of the 2011 and 2018 usual interstitial pneumonia (UIP) criteria according to the joint guidelines from international thoracic and respiratory societies. Material and methodsTwo observers independently evaluated 126 HRCT, with examples of most common parenchymal patterns, and noted the presence of each pattern. The readers also noted whether the findings met the 2011 criteria for UIP. In a second reading, the same readers noted whether the HRCT met the UIP criteria according to the 2018 UIP update. ResultsThe kappa values for interobserver variability for the different patterns ranged from 0.28 (intralobular lines) to 0.85 (tree-in-bud nodules). The kappa value for UIP pattern was similar for 2011 and 2018 criteria, 0.58 and 0.69, respectively. Compared to the 2011 UIP criteria, there was no statistically significant difference in the number of HRCT classified as UIP using the 2018 criteria. ConclusionsThere is a substantial variation in interobserver agreement between the different parenchymal patterns, which suggests that some patterns a more easily identified than others. There is also a considerable reader variation in the assessment of UIP applying the 2011 UIP criteria as well as applying the 2018 UIP update.

Reproducibility of Grading in Chronic Intervillositis of Unknown Etiology.

Chronic intervillositis of unknown etiology (CIUE) is a rare placental inflammatory process associated with pregnancy loss and recurrence. We conducted a quality assurance study to assess the diagnostic accuracy and reproducibility of CIUE grading at our institution. Hematoxylin and eosin-stained slides from 20 CIUE cases (31 slides) were reviewed by 7 perinatal pathologists in 2 sequential rounds. Reviewers were instructed to use the diagnostic criteria they were presently following for CIUE and to grade each slide according to the Rota scheme. In the first round, 20 slides were assessed. The diagnostic accuracy was 94%, the average percent agreement of Rota grade was 79%, and the Fleiss' kappa value for interobserver variability was 0.54. The results were reviewed by all pathologists with diagnostic and grading criteria agreed upon prior to the second round. In round 2, the remaining 11 slides were assessed. Diagnostic accuracy was 83%, the average percent agreement on Rota grade was 70%, and the Fleiss' kappa value for interobserver variability was 0.36. Overall, diagnostic accuracy was high and agreement on Rota grade was moderate. Group review did not appear to improve accuracy. Simplifying CIUE grading to a low-grade/high-grade scheme (<50% or ≥50%) might improve grading reproducibility.

EP1.09-03 Interobserver Variability Study of PD-L1 Immunostaining in Non-Small Cell Lung Cancer

Pathologic assessment of programmed cell death ligand 1 (PD-L1) immunohistochemical staining now have been used for treatment of non-small cerll lung cancer in advanced stage. But, due to misinterpretation or unspecified staining pattern, interpretation of PD-L1 staining may be difficult. This study was to investigate interobserver variability between three different pathologists to see if the staining can be safely assessed in practice. PD-L1 immunostaining (SP263 and SP142) was performed in representative sections of 70 adenocarcinoma cases and 58 squamous cell carcinoma cases which surgically resected at the chonnam National University Hwasun Hospital, Republic of Korea between 2013 to 2015. Three pathologists individually evaluated the percentage of positive tumor cells, scoring each sample applying cutoff levels used in clinical practice: <1% positive tumor cells (score 0), 1–4% (score 1), 5–9% (score 2), 10–49% (score 3), and >50% positive tumor cells (score 4). Three pathologists were in agreement in 121 of 128 (94.5%) of scored cases in the SP263 test and 109 of 128 (85.2%) in the SP142 test. In 26 remaining cases, at least two pathologists were in aggrement of scored cases. Weighted kappa value for interobserver variability between pathologists was respectively 0.892 in the SP263 test and 0.837 in the SP142 test. The number of differently classified cases was significantly higher for the ≥1% cutoff value than every other cutoff value. Also, a significantly better agreement between pathologists was seen using ≥50% as cutoff. No statistically significant differences were seen between adenocarcinoma and squamous cell carcinoma. Interobserver agreement remains an important challenge, and the ≥1% cutoff value seems to be problematic. But, this may be overcome by training, longer experience and optimization of cutoff value.

Evaluation of an abbreviated screening MRI protocol for patients at risk for hepatocellular carcinoma

In this study, we compare an abbreviated screening MRI protocol (aMRI), utilizing only dynamic contrast-enhanced images, to a conventional liver MRI (cMRI) for the characterization of observations in at-risk patients. 164 consecutive HCC screening MRIs were retrospectively analyzed. Two sets of de-identified image sets were created: one with all acquired sequences including T2- and diffusion-weighted sequences (cMRI), and one with only T1-weighted precontrast and dynamic post-contrast images utilizing an extracellular gadolinium contrast agent (aMRI). Three readers assigned a LI-RADS score based on the lesion with the highest LI-RADS category using the aMRI and cMRI datasets during separate reads. There was no change between the aMRI and cMRI LI-RADS categorization in 93%, 96%, and 96% of cases for readers 1, 2, and 3, respectively. In the majority of the discrepant cases, the score increased from LI-RADS 3 to LI-RADS 4 due to the presence of ancillary features on T2 and DWI. Kappa values for interobserver variability demonstrated fair-to-moderate LI-RADS agreement among the 3 readers. There was strong agreement between the abbreviated T1-only MRI protocol and a full liver MRI, with only 5% of cases changing LI-RADS categorization due to the inclusion of T2 and DWI. The estimated time to run this abbreviated MRI is approximately 7-10min, possibly allowing for a more cost-effective screening MRI than our cMRIs.

Epiphyseal stress fractures of finger phalanges in adolescent climbing athletes: a 3.0-Tesla magnetic resonance imaging evaluation

To study the value of 3.0-Tesla magnetic resonance imaging for baseline and follow-up assessment of epiphyseal finger phalanx stress fractures in a collective of 7 consecutive adolescent climbing athletes. Baseline MRI was performed in 8 fingers of 7 adolescent athletes (mean age 13.8years, female:male = 2:5) with clinically suspected stress fracture of the fingers acquired during climbing sports. Follow-up MRI was performed after functional therapy with training interruption for 6weeks (n = 6) and 12weeks (n = 1). Images were analysed retrospectively and independently by two readers using an MRI grading score from 0 (no pathology) to 4 (bone marrow oedema and clear depiction of a sharp fracture line with surrounding inflammatory soft tissue reaction). A total of 8 baseline and 7 follow-up MRIs were analysed. In 7 out of 8 fingers a stress fracture line Salter-Harris III and in all fingers a bone marrow oedema were diagnosed at the epiphyseal base of the middle phalanx. The average grading score was 3.37 in the initial MRI and 1.43 in the follow-up MRI indicating fracture healing in all fingers. Kappa value for interobserver variability was 0.86, representing almost perfect interobserver agreement. 3-T MRI is a promising diagnostic technique for baseline assessment of epiphyseal finger phalanx stress fractures and for follow-up evaluation of fracture healing.

Feasibility of abdominal plain film images in evaluation suspected drug smuggler

ObjectiveDrug smuggling in the gastrointestinal tract has soared within the last 20 years. Though illegal substances in the gastrointestinal tract can be visualized with ultrasound, MRI and CT, the abdominal radiograph has by far remained the most frequently used way of detecting smuggled drugs. The purpose of the study was to evaluate the inter-radiologist interpretation error and the reliability of the abdominal radiograph in detecting smuggled drugs. Materials and methodsA total of 279 abdominal radiographs of suspected smugglers were classified by three radiologists as clearly positive or negative for drug smuggling. All available information about the cases was collected from the customs officers and police. ResultsOut of these cases 203 (73%) were interpreted as negative and 35 (13%) as positive by all three radiologists. In 86% of the cases there was, therefore, an inter-radiological agreement in interpreting the images. In 41 (14%) cases, however, there was an inter-radiologist disagreement. Kappa-value for inter-observer variability was 0.70. ConclusionsIn up to a seventh of the abdominal radiographs the interpretation can be challenging even for an experienced radiologist. False positive interpretation can lead to innocent passengers being detained in vain. As negatively interpreted images usually result in releasing of the suspect, there is no way of knowing how many false negative occur. This makes the abdominal radiograph a suboptimal examination, and low dose CT should be considered as the screening modality for gastrointestinal drug smugglers.

Comparison of 1.5T and 3T MRI scanners in evaluation of acute bone stress in the foot

BackgroundBone stress injuries are common in athletes and military recruits. Only a minority of bone stress changes are available on plain radiographs. Acute bone stress is often visible on MRI as bone marrow edema, which is also seen in many other disease processes such as malignancies, inflammatory conditions and infections. The purpose of this study was to investigate the ability of radiographs, 1.5T and 3T MRI to identify acute bone marrow changes in the foot.MethodsTen patients with 12 stress fractures seen on plain radiographs underwent MRI using 1.5T and 3T scanners. T1 FSE and STIR axial, sagittal, and coronal view sequences were obtained. Two musculoskeletal radiologists interpreted the images independently and by consensus in case of disagreement.ResultsOf the 63 acute bone stress changes seen on 3T images, 61 were also seen on 1.5T images. The sensitivity of 1.5T MRI was 97% (95% CI: 89%-99%) compared with 3T. The 3T MRI images where, therefore, at least equally sensitive to 1.5T scanners in detection of bone marrow edema. On T1-weighted sequences, 3T images were slightly superior to 1.5T images in visualizing the demarcation of the edema and bone trabeculae. The kappa-value for inter-observer variability was 0.86 in the MRI indicating substantial interobserver agreement.ConclusionsOwing to slightly better resolution of 3T images, edema characterization is easier, which might aid in the differential diagnosis of the bone marrow edema. There was, however, no noteworthy difference in the sensitivity of the 1.5T and 3T images to bone marrow edema. Routine identification of acute bone stress changes and suspected stress injuries can, therefore, be made with 1.5T field strength.

Characterization of colorectal tumors using narrow-band imaging magnification: combined diagnosis with both pit pattern and microvessel features

Objective. We aimed to clarify the clinical usefulness of narrow-band imaging (NBI) magnification for evaluating both pit appearances and microvessel architecture in comparison with evaluation of microvessel architecture alone in invasion depth of colorectal tumors. Material and methods. A total of 189 colorectal lesions [37 adenomas, 73 intramucosal to scanty submucosal invasive carcinomas and 79 massive submucosal invasive (SM-m) carcinomas] were analyzed. All lesions showing irregular pit structure were observed by NBI magnifying endoscopy. Based on both pit appearance and microvessel features, lesions were classified into three grades (C1, C2, C3), as described previously. Also, lesions were classified as high or low by microvessel irregularity. Furthermore, the histopathological background and the inter- and intraobserver variability of C subtype were assessed. Results. The SM-m rate of C1, C2 and C3 was 5.2, 60.5 and 92.7%, respectively. On the other hand, SM-m rate of lesion with mildly and highly irregular vessels was 13.3 and 55.0%, respectively. For the histopathological background, a high percentage of destruction of the glandular orifice, disappearance of the lamina muscularis mucosae and superficial exposure of desmoplastic reaction (DR) were observed in type C3. For C subtype classification of NBI magnification findings, the kappa value for interobserver variability was 0.749, and the kappa value for intraobserver variability was 0.745. Conclusions. NBI magnification findings evaluated by both pit appearances and microvessel features (C subtype) showing a good kappa value in variability are more useful in invasion depth diagnosis of colorectal tumor than those evaluated by microvessel features alone.

Intraobserver and interobserver variability and spatial differences in histologic examination of carotid endarterectomy specimens

Studies using histologic examination and protein analysis of atherosclerotic plaques are increasingly being performed, but reproducibility of plaque histology and variation of plaque composition among different parts of the plaque, which are key to reliability of these studies, are relatively unexplored. Therefore, this study investigated the intraobserver and interobserver variability of plaque histology and spatial variability in plaque composition. Atherosclerotic plaques (n = 100) obtained during carotid endarterectomy were divided into 0.5-cm segments. Paraffin sections were stained and semiquantitatively analyzed (four categories: no, minor, moderate, and heavy) for fat, macrophages, smooth muscle cells, collagen, calcification, thrombus, and overall phenotype. First, to determine the intraobserver and interobserver reproducibility, two independent observers independently analyzed the plaques. Second, to investigate spatial variability in plaque composition, histologic appearances of the culprit lesions (0-segment) were compared with the histologic appearances of adjacent (+5 mm) and more distant (+10 mm) plaque segments of 30 specimens. The kappa values for intraobserver variability of fat, macrophages, smooth muscle cells, collagen, calcifications, thrombus, and overall phenotype were 0.83, 0.85, 0.71, 0.63, 0.81, 0.80, and 0.86, respectively, and kappa values for interobserver variability were 0.68, 0.74, 0.54, 0.59, 0.82, 0.75, and 0.71, respectively. Comparison of the histologic scorings of adjacent segments revealed a mean kappa of 0.40 (range, 0.33 to 0.60). When the culprit segment was compared with the more distant segment, the mean kappa was 0.24; however, in 91% of cases, the difference between the culprit segment and the distal segment was one category or less. Semiquantitative analysis of carotid atherosclerotic plaque histology was well reproducible, both intraobserver and interobserver. Although variation between different plaque segments in histologic appearance was observed, differences were small in almost all cases. Variability in histologic examination needs to be taken into account in studies comparing plaque imaging with histopathology and plaque research studies.

Diagnostic accuracy of portal-phase CT and MRI with mangafodipir trisodium in detecting liver metastases from colorectal carcinoma

To compare the diagnostic accuracy of single section spiral computed tomography (CT) and magnetic resonance imaging (MRI) with tissue-specific contrast agent mangafodipir trisodium (MnDPDP) in the detection of colorectal liver metastases. One hundred and twenty-five consecutive patients undergoing surgery for primary and/or metastatic disease were evaluated using CT (5 mm collimation and reconstruction interval, pitch 2), two-dimensional fast spoiled gradient echo (2D FSPGR) T1 and single shot fast-spin echo (SSFSE) T2 weighted breath-hold MRI sequences, performed before and after intravenous administration of MnDPDP. The reference standards were intraoperative ultrasound and histology. The per-patient accuracy of CT was 72.8 versus 78.4% for unenhanced MRI (p = 0.071) and 82.4% for MnDPDP-enhanced MRI (p = 0.005). MnDPDP-enhanced MRI appeared to be more accurate than unenhanced MRI but this was not significant (p = 0.059). The sensitivity of CT was 48.4% versus 58.1% for unenhanced MRI (p = 0.083) and 66.1% for MnDPDP-enhanced MRI (p = 0.004). The difference in specificity between procedures was not significant. The per-lesion sensitivity was 71.7, 74.9 and 82.7% for CT, unenhanced MRI, and MnDPDP-enhanced MRI, respectively; the positive predictive value of the procedures was respectively 84.0, 96.0 and 95.8%. MnDPDP-enhanced MRI provided a high level diagnostic confidence in 92.5% of the cases versus 82.5% for both unenhanced MRI and CT. The kappa value for inter-observer variability was >0.75 for all procedures. The diagnostic accuracy and sensitivity of MnDPDP-enhanced MRI is significantly higher than single section spiral CT in the detection of colorectal cancer liver metastases; no significant difference in diagnostic accuracy was observed between unenhanced MRI and MnDPDP-enhanced MRI.

Imaging of Patellar Cartilage with a 2D Multiple-Echo Data Image Combination Sequence

We sought to evaluate the diagnostic value of a 2D multiple-echo data image combination (MEDIC) MRI sequence in the detection of patellar cartilage defects. Our study included 52 consecutive patients who had knee surgery within 4 months of undergoing an MRI examination including an axial 2D MEDIC (TR/TE, 884/26; flip angle, 30 degrees ) sequence. Cartilage was surgically graded on a 5-point scale: 0, normal; 1, softening or swelling; 2, partial thickness defect; 3, fissuring to the level of the subchondral bone; or 4, exposed subchondral bone. Cartilage was graded on MRI according to a scale that was almost identical to the surgical scale except that grade 1 lesions were defined as signal alteration or swelling of cartilage. Two blinded reviewers independently analyzed patellar cartilage. Sensitivity, specificity, accuracy, and weighted kappa values for interobserver variability were calculated. Low-grade cartilage lesions predominated in our study group. When grade 2 or higher was considered the threshold for relevance, the sensitivity, specificity, and accuracy for the MEDIC sequence was as high as 79%, 82%, and 81%, respectively. Increasing the threshold of relevance to grade 3 increased the sensitivity, specificity, and accuracy to as high as 83%, 91%, and 90%, respectively. Interobserver agreement for the MEDIC sequence was good (weighted kappa = 0.68). The 2D MEDIC sequence performs comparably to previously described sequences optimized for cartilage imaging such as the 3D double-echo steady-state or 3D spoiled gradient-recalled sequences with good interobserver agreement, high sensitivity, and excellent specificity for revealing low- to intermediate-degree cartilage defects.

Ultrasonographic screening for developmental dysplasia of the hip in infants

We studied the reproducibility of ultrasonographic screening examination of the hip when read by diagnostic radiographers. In order to determine interobserver variability, 200 ultrasonograms were classified according to Graf's method by five observers (four radiographers and one radiologist). The kappa values for interobserver variability indicated moderate agreement (kappa 0.47) for the exact Graf classification and substantial agreement (kappa 0.65) for the classification of normal (type I) versus abnormal (type IIa-IV). Agreement was significantly different for normal, immature and abnormal hips. Comparison of the findings in our interobserver study with existing information based on other examinations and treatment revealed that only a small number of infants with mildly dysplastic hips would have been typed as normal by some observers as a result of observer variability. In conclusion, the interobserver agreement on the ultrasound assessment of the hip was good enough for screening purposes. Observer variability did not result in any severe cases being missed.

Focal radial styloid abnormality as a manifestation of de Quervain tenosynovitis.

de Quervain disease is a stenosing tenosynovitis of the first dorsal wrist compartment. The purpose of this study was to determine whether focal radial styloid abnormality (cortical erosion, sclerosis, or periosteal bone apposition) as shown by radiography can be an indicator of de Quervain tenosynovitis. A retrospective review of 49 radiographs from 45 patients in whom the clinical diagnosis of de Quervain tenosynovitis was confirmed (positive findings on Finkelstein's test) and 64 radiographs from 62 asymptomatic patients was carried out independently by two musculoskeletal radiologists in a blinded fashion. Findings on radiographs were assessed for focal radial styloid abnormality and assigned a diagnostic grade (1, definitely normal; 2, probably normal; 3, equivocal; 4, probably abnormal; 5, definitely abnormal). Receiver operating characteristic curves were constructed and compared. Kappa statistics for interobserver and intraobserver variability were calculated. The presence of focal radial styloid abnormality correlated significantly with the presence of de Quervain tenosynovitis (p < 0.05). The areas under the receiver operating characteristic curves for each reviewer equaled 0.71 and 0.76. Kappa values for interobserver variability equaled 0.44 (moderate agreement), and intraobserver variability equaled 0.62 (substantial agreement). Focal radial styloid abnormality is an indicator of de Quervain stenosing tenosynovitis of the wrist.

MR imaging of anterior cruciate ligament reconstruction graft.

The objective was to determine the MR imaging findings that differentiate intact anterior cruciate ligament reconstruction graft, partial-thickness tear, and full-thickness tear, using arthroscopy as the gold standard. Sixteen consecutive MR imaging examinations were retrospectively and independently evaluated by two musculoskeletal radiologists for primary signs (graft signal, orientation, fiber continuity, complete discontinuity, and thickness) and secondary signs (anterior tibial translation, uncovered posterior horn lateral meniscus, posterior cruciate ligament hyperbuckling, and abnormal posterior cruciate ligament line) of anterior cruciate ligament reconstruction graft tear in 15 patients with follow-up arthroscopy. Results were compared with arthroscopy, and both receiver operating characteristic curves and kappa values for interobserver variability were calculated. Arthroscopy revealed four full-thickness graft tears, seven partial-thickness tears, and five intact grafts. Of the primary signs, graft fiber continuity in the coronal plane and 100% graft thickness in the sagittal or coronal plane were most valuable in excluding full-thickness tear. Complete discontinuous graft in the coronal plane also was valuable in diagnosis of full-thickness tear. Of the secondary signs, anterior tibial translation and uncovered posterior horn lateral meniscus assisted in differentiating graft tear (partial or full thickness) from intact graft. The other primary and secondary signs were less valuable. Kappa values were highest for graft fiber continuity and graft discontinuity in the coronal plane. Full-thickness anterior cruciate ligament graft tear can be differentiated from partial-thickness tear or intact graft by evaluating for graft fiber continuity (coronal plane), complete graft discontinuity (coronal plane), and graft thickness (coronal or sagittal plane).