Radiographic Landmarks Research Articles

Clinical benchmark dataset for AI accuracy analysis: quantifying radiographic annotation of pelvic tilt

Radiographic landmark annotation determines patients’ anatomical parameters and influences diagnoses. However, challenges arise from ambiguous region-based definitions, human error, and image quality variations, potentially compromising patient care. Additionally, AI landmark localization often presents its predictions in a probability-based heatmap format, which lacks a corresponding clinical standard for accuracy validation. This Data Descriptor presents a clinical benchmark dataset for pelvic tilt landmarks, gathered through a probabilistic approach to measure annotation accuracy within clinical environments. A retrospective analysis of 115 pelvic sagittal radiographs was conducted for annotating pelvic tilt parameters by five annotators, revealing landmark cloud sizes of 6.04 mm-17.90 mm at a 95% dataset threshold, corresponding to 9.51°–16.55° maximum angular disagreement in clinical settings. The outcome provides a quantified point cloud dataset for each landmark corresponding to different probabilities, which enables assessment of directional annotation distribution and parameter-wise impact, providing clinical benchmarks. The data is readily reusable for AI studies analyzing the same landmarks, and the method can be easily replicated for establishing clinical accuracy benchmarks of other landmarks.

Radiographically Apparent Acetabular Sourcil Landmarks Are Created by Comparable Regions of the Pelvis With Extraarticular Bone Variably Confounding Estimates of Joint Coverage.

The severity of hip dysplasia is characterized by radiographic measurements that require user definition of the acetabular sourcil edge, a bony landmark for which the corresponding three-dimensional (3D) anatomy is not well defined in any imaging plane. To use digitally reconstructed radiographs to determine: (1) What 3D anatomy is contributing to the "acetabular sourcil" location used to make lateral center-edge angle (LCEA) and anterior center-edge angle (ACEA) measurements in standing AP and false-profile radiographic views, respectively? (2) How do intraobserver and interobserver agreement in LCEA and ACEA translate into agreement of the 3D anatomy being evaluated? (3) How distinct are regions around the acetabular rim circumference that are evaluated by LCEA and ACEA measurements on radiographs? Between January 2018 and May 2019, 72 patients were indicated for periacetabular osteotomy to treat hip dysplasia or acetabular retroversion at our institution. From these patients, a series of 10 patients were identified of the first 12 patients in 2018 who were treated with periacetabular osteotomy, excluding two with missing or low-quality clinical imaging. A second series of 10 patients was identified of the first 11 patients in 2019 who were treated with periacetabular osteotomy and concurrent hip arthroscopy, excluding one who was missing clinical imaging. Pelvis and femoral bone surface models were generated from CT scans of these two series of 10 patients. There were 15 female and five male patients, with a median patient age of 18 years (IQR 17 to 23 years), a preoperative LCEA of 22° (IQR 18° to 24°), and a preoperative ACEA of 23° (IQR 18° to 27°). Exclusion criteria included missing preoperative CT or standard clinical radiographic imaging or severe joint incongruity. To address our first study question, digitally reconstructed radiographs matching each patient's standing AP and false-profile clinical radiographs were created from the segmented CT volumes. A board-certified orthopaedic surgeon and three trained researchers measured LCEA and ACEA on the digitally reconstructed radiographs, and the selected sourcil points were projected back into coordinates in the 3D anatomic space. To address our second study question, intraobserver and interobserver agreement in radiographic coverage angles were related to variations in 3D coordinates of the selected bony anatomy. Lastly, to address our third study question, 3D locations around the acetabular rim identified as contributing to the lateral and anterior sourcil points were summarized across patients in a clockface coordinate system, and statistical analysis of the "time" separating the 3D acetabular contributions of the sourcil edges was performed. The 3D anatomy contributing to the lateral sourcil was a variable length (27 mm [IQR 15 to 34 mm]) span of the laterosuperior acetabular edges, with contributions by the anterior inferior iliac spine in 35% (7 of 20) of hips. The anterior sourcil reflected a 28-mm (IQR 25 to 31 mm) span of bone from the medial ilium (posterior-medial to the anterior-inferior iliac spine and anterior-lateral to the arcuate line) to the anterior and lateral edges of the acetabulum. Interobserver variability was good for LCEA (intraclass correlation coefficient [ICC] 0.82 to 0.83) and moderate to good for ACEA (ICC 0.73 to 0.79), whereas the agreement in identified 3D sourcil locations varied widely (ICC 0.32 to 0.95). The acetabular edge of the 3D anatomy contributing to the anterior sourcil overlapped the circumferential range of the acetabular rim contributing to the lateral sourcil. Projection of two-dimensional radiographic landmarks contributing to the diagnosis of structural hip deformity into 3D allowed for the identification of the overlapping bony anatomy contributing to radiographically visible anterior and lateral sourcil edges. This work leveraging digitally reconstructed radiographs and 3D pelvis anatomy has found that bone outside the joint contributes to the radiographic appearance of the sourcil and may variably confound estimates of joint coverage. Furthermore, the substantial overlap between the acetabular bone contributing to measurement of the LCEA and ACEA would indicate that these angles measure similar acetabular deformity, and that additional measures are needed to assess anterior coverage independent of lateral coverage.

Poster 246: Radiographic Landmarks for Femoral Tunnel Positioning During MPFC Reconstruction Vary With Trochlear Dysplasia

Objectives: Accurate femoral tunnel positioning during medial patellofemoral complex (MPFC) reconstruction has been shown to be critical in optimizing outcomes and avoiding complications. Multiple anatomic and radiographic studies have described methods to accurately identify the appropriate femoral attachment of the MPFC, commonly referencing the adductor tubercle and medial epicondyle as landmarks. However, these studies have been performed in normal knees, and the relevance to knees with patellar instability and anatomic risk factors has not been described. Therefore, the purpose of this study was to compare the radiographic positions of commonly utilized landmarks between symptomatic and normal knees. Methods: Three-dimensional (3D) models were created from computed tomography scans of knees with patellar instability and compared to age- and sex-matched control knees. On the 3D models, the locations of the adductor tubercle (AT), medial epicondyle (ME) and gastrocnemius tubercle (GT) were marked. A 2-dimensional perfect lateral view was then created from these models to simulate radiographs, and the radiographic location of each landmark was described with respect to the following: 1) anteroposterior relation to the posterior cortical line and 2) proximal-distal relationship to the posterior condylar line. The position of each landmark was compared between symptomatic and control groups, and linear regression analysis was performed to assess variations in position with severity of anatomic risk factors including trochlear dysplasia, tuberosity lateralization and patella alta. Results: Forty knees were included in this study, including 20 symptomatic and 20 control knees (8 men and 12 women in each group, age 21.5+/-6.9 years old vs 21.3+/-7.0 years old). On the 2D (radiographic) views, the ME landmarks in the symptomatic group were found to be significantly more posterior (6.9+/-3.7mm vs 9.5+/-4.2 mm, p = 0.031) and significantly more distal (9.2+/-4.4 vs 12.1+/-5.6, p = 0.047) than in the control group. For the AT, no significant differences were noted in the anteroposterior (0.7+/-2.4 vs 1.3+/-2.6 mm anterior to the posterior cortex, p = 0.238) or proximal-distal (4.9+/-2.8 vs 5.4+/-2.3mm proximal to the posterior condylar line, p = 0.259) positions. Similarly, no differences for the GT were found (4.4+/-3.4 vs 2.8+/-3.7 mm posterior to the posterior cortex, p = 0.100; and 5.1+/-4.6 vs 4.0+/-3.2 mm distal to the posterior condylar line, p = 0.175). An association between severity of trochlear dysplasia and posterior position of the AT was identified (R = 0.43, R2 = 0.18, p = 0.058). Conclusions: Although intraoperative fluoroscopy is commonly utilized to identify femoral tunnel positioning during MPFC reconstruction, this study demonstrates that commonly utilized anatomic reference points such as the medial epicondyle in symptomatic knees are more posterior and more distal on perfect lateral radiographs when compared to normal knees. Furthermore, the severity of trochlear dysplasia appears to influence radiographic AT position. These findings should be taken into consideration when utilizing fluoroscopy during MPFC reconstruction. Further studies are needed to identify optimal femoral tunnel positioning for knees with anatomic risk factors when performing surgical treatment for patellar instability.

The Sensitivity of the Scapholunate Interval and Bony Landmarks to Wrist Rotation on Posteroanterior Radiographs.

The objective of this study was to examine the effect of wrist rotation on the scapholunate interval in the posteroanterior radiograph and to identify radiographic landmarks on the posteroanterior projection that can be used to assess position. Eleven healthy cadaveric wrists were radiographed in the neutral position and subsequently were rotated and imaged from 30° pronation to 30° supination in 10° intervals. At each interval, the scapholunate interval was measured as well as the following landmarks: (1) the visible perimeter of the base of the hook of the hamate; (2) the radial-ulnar distribution of the dorsal nonarticular surface of the distal third metacarpal head; (3) the radial-ulnar distribution of the pisiform about the longitudinal axis of the ulna; and (4) the overlap of the pisiform and triquetrum. The scapholunate interval was largest in the neutral position and linearly decreased by 34% for every 10° of pronation and decreased nonlinearly by 86% after the first 10° of supination. The appearance of the distal third metacarpal head was shown to be sensitive to both pronation and supination. The perimeter of the hook of the hamate and the distribution of the pisiform compared to the ulna were both shown to be sensitive to supination, whereas overlap of the pisiform and triquetrum was not shown to be sensitive to either direction of rotation. Our results highlight the significant effect of rotation on radiographic landmarks at the wrist, indicating that 10° of supination can drastically alter the developed radiograph.

Are Variable Screw Angle Change and Screw-to-Vertebral Body Ratio Associated With Radiographic Subsidence Following Anterior Cervical Discectomy and Fusion?

Retrospective Cohort. This study aims to assess the easily measurable radiographic landmarks of screw-to-vertebral body ratio and changes in screw angle to identify if they are associated with early subsidence following an Anterior cervical discectomy and fusion (ACDF). A retrospective cohort study was conducted on patients undergoing 1-3 level ACDF with allograft or PEEK cages. Preoperative, immediate postoperative, and 6-month postoperative radiographs were analyzed to measure intradiscal height (or distance between 2 vertebral bodies) as an anterior vertebral distance (AVD), middle (MVD), and posterior (PVD), screw angle, screw-to-vertebral body length ratio, and interscrew distance. Multivariate stepwise regression analyses were performed. 92 patients were included (42 single-level, 32 two-level, and 18 3-level ACDFs). In single-level ACDFs, a decrease in the caudal screw angle was associated with a decrease in AVD (=.001) and MVD (P = .03). A decrease in the PVD was associated with a decrease in segmental lordosis (P < .001). For two-level ACDFs, a higher caudal screw-to-body ratio was associated with a lower MVD (P = .01). Six months following an ACDF for degenerative pathology, a decrease in the caudal screw angle was associated with an increase in radiographic subsidence at the antero-medial aspect of the disc space albeit largely subclinical. This suggests that the caudal screw angle change may serve as a reliable radiographic marker for early radiographic subsidence. Furthermore, a greater screw-to-vertebral body ratio may be protective against radiographic subsidence in two-level ACDF procedures.

Automated radiographic hip morphology measurements: An open-access method

ObjectiveThe aim of this study is to present a newly developed automated method to determine radiographic measurements of hip morphology on dual-energy x-ray absorptiometry (DXA) images. The secondary aim was to compare the performance of the automated and manual measurements. Design30 DXA scans from 13-year-olds of the prospective population-based cohort study Generation R were randomly selected. The hip shape was outlined automatically using radiographic landmarks from which the acetabular depth-width ratio (ADR), acetabular index (AI), alpha angle (AA), Wiberg and lateral center edge angle (WCEA) (LCEA), extrusion index (EI), neck-shaft angle (NSA), and the triangular index (TI) were determined. Manual assessments were performed twice by two orthopedic surgeons. The agreement within and between observers and methods was visualized using Bland-Altman plots, and the reliability was studied using the intraclass correlation coefficient (ICC) with 95 % confidence intervals (CI). ResultsThe automated method was able to perform all radiographic hip morphology measurements. The intermethod reliability between the automated and manual measurements ranged from 0.57 to 0.96 and was comparable to or better than the manual interobserver reliability, except for the AI. ConclusionThis open-access, automated method allows fast and reproducible calculation of radiographic measurements of hip morphology on right hip DXA images. It is a promising tool for performing automated radiographic measurements of hip morphology in large population studies and clinical practice.

Fluoroscopy-Assisted Computer Navigation Accurately Determines Cup Position and Leg Length for Anterior Hip Arthroplasty.

Recently, fluoroscopy-assisted computer navigation has been developed to assess intraoperative cup inclination/anteversion and leg-length discrepancy (LLD) in the operating room. However, there is a relative dearth of studies investigating the accuracy of this software compared with postoperative radiographs. We prospectively enrolled 211 navigated anterior total hip arthroplasties using fluoroscopy-assisted computer navigation software. Intraoperative navigated measurements were compared with postoperative anteroposterior radiographs to assess accuracy of cup inclination/anteversion and LLD. Continuous variables were analyzed using the Student's t test, and categorical variables were analyzed using Fisher's exact test. On postoperative radiographs, 94.3% of cups (199 of 211) were positioned within the Lewinnek "safe zone," compared with 99.1% navigated intraoperatively (P=.01). Eighty-two percent of hips (174 of 211) were navigated intraoperatively to LLDs within ±2 mm; on postoperative radiographs, 65% of hips (138 of 211) had LLDs within ±2 mm (P=.0001). Intraoperatively, 100% of hips (211 of 211) were navigated to LLDs within ±5 mm; similarly, on postoperative radiographs, 98% of hips (207 of 211) had LLDs within ±5 mm (P=.12). A novel fluoroscopy-assisted computer navigation platform accurately assessed intraoperative cup position and LLD during anterior total hip arthroplasty. Careful attention to fluoroscopic technique, positioning of radiographic landmarks, and knowledge of the limitations of fluoroscopy, including parallax effect, are important concepts that surgeons should incorporate into their decision algorithm. [Orthopedics. 2024;47(4):e174-e180.].

COMPARISON OF THE ACCURACY OF MANUAL CEPHALOMETRY RADIOGRAPHIC MEASUREMENTS AND WEB-BASED DIGITAL SOFTWARE

Background: Determining anatomical landmarks - in the cranium, maxilla and mandible - as well as measuring skeletal and dental angles on cephalometric radiographs are supporting examinations that determine important diagnoses in orthodontic treatment. Traditionally, cephalometric analysis has been performed by tracing radiographic landmarks on acetate overlays and measuring linear and angular variables using protractor. However, despite its widespread use in orthodontics, the technique is time consuming and has several drawbacks, including a high risk of error in tracing, landmark identification, and measurement. Objective: to evaluate the difference of cephalometric measurements using manual and digital technique. Method: pre-treatment cephalometric digital radiographs of 40 patients were traced manually and digitally using WebCeph Ver. 1.0.0 computer software program by the same investigator. A total of 8 anatomical landmarks were located and five angular measurements based on Steiner Analysis were measured. Independent t-tests and Mann-Whitney tests were used to compare the difference of manual and digital measurements. Result: the p-values for SNA, SNB, ANB, I-NA, I-NB were greater than 0.05 (p&gt;0.05). Conclusion: There were no significant difference between manual and digital tracing cephalometric technique using WebCeph for SNA, SNB, ANB, I-NA and I-NB Keywords: Cephalometric measurement, Digital cephalometric analysis, Steiner Analysis, Tracing cephalometry, WebCeph

Radiographic Landmark Measurements for the Femoral Footprint of the Medial Patellofemoral Complex May Be Affected by Visible Femoral Shaft Length on Lateral Knee Radiographs

To assess the effect of visible femoral shaft length on the accuracy of radiographic landmarks of the medial patellofemoral complex (MPFC). In 9 cadaveric knees, the MPFC footprint was exposed on the medial femur, and its proximal and distal boundaries were marked. Lateral fluoroscopic images of the knee were assessed in 1-cm length increments, beginning 1 cm proximal to the medial condyle and continuing proximally to 8 cm. The MPFC midpoint was described on each image relative to the posterior cortical line of the femur and a line perpendicular to this line through the proximal margin of the medial condyle. In addition, the MPFC midpoint was assessed relative to a line from the proximal posterior cortex to the midpoint of Blumensaat line. Using the posterior cortical line as a reference, the MPFC radiographic landmark moved anteriorly with decreasing visible femoral shaft on radiographs, particularly at 4 cm and less. However, no proximal-distal change was noted. Linear regression analysis demonstrated a relationship between visible femoral shaft and MPFC position on radiographs (R= 0.461, R2= 0.212, B= -0.636, P < .001). For every centimeter decrease in the visible femoral shaft, the radiographic MPFC footprint moved anteriorly by 0.636 mm. Receiver operating characteristic curve analysis revealed that a minimum of 4 cm of femoral shaft on lateral radiographs is required for accurate MPFC footprint localization (area under the curve= 0.80; sensitivity= 76.7%; specificity= 69.0%; P < .001). In contrast, no anterior-posterior change was seen when referencing a line from the proximal posterior cortex to the midpoint of Blumensaat line. When using the posterior cortical line to identify the midpoint of the MPFC, at least 4 cm of femoral shaft should be visible for accurate assessment. If less than 4 cm of shaft is visible, a line through the midpoint of Blumensaat line and the proximal posterior cortex can be used as an alternative method to estimate the position of the femoral footprint. As fluoroscopy is frequently used intraoperatively for MPFC reconstruction, our findings may serve as a guide when assessing femoral tunnel placement on fluoroscopy.

The distal tibiofibular syndesmosis is a reliable landmark for 3° varus tibial resection in total knee arthroplasty: a radiological evaluation on 1296 cases.

The purpose of this study was to evaluate the reproducibility and the accuracy of distal tibiofibular syndesmosis (DTFS) as landmark to perform controlled varus tibial resections during total knee arthroplasty (TKA). The hypothesis was that DTFS can be used to perform an accurate 3° varus tibial cut. A retrospective analysis on a consecutive series of standard weightbearing full-length anteroposterior views of the lower limbs radiographic images was conducted. For each radiograph, the hip-knee-ankle (HKA) angle, the angle between the tibial mechanical axis and the line connecting the centre of the tibial spines and the DTFS (tibiofibular angle, TFA) and the medial proximal tibial angle (MPTA) were calculated. Each measurement was carried out twice by three independent observers, and intra- and inter-observer measurement reliability were assessed using the intraclass correlation coefficient (ICC) analysis. A total of 1296 lower limbs were analysed from a series of 648 weightbearing full-length anteroposterior radiographs. The ICC were > 90% for all measurements. The mean TFA value was 2.94 ± 0.68 (range 2.38-3.51). No differences were detected comparing the mean TFA value on the right and left limb (p = 0.795) as well as comparing the values in male and female patients (p = 0.691). Linear regression analysis did not find statistically significant correlation between TFA and MPTA, or TFA and HKA angles, respectively. The distal tibiofibular syndesmosis is a reliable and easy reproducible radiographic landmark that can be used when planning a 3° varus tibial cut. Future studies are needed to confirm the validity of this method also in clinical settings. IV, retrospective case series.

The Potential Usefulness of ChatGPT in Oral and Maxillofacial Radiology.

Aim This study aimed to evaluate the potential usefulness of Chat Generated Pre-Trained Transformer-3 (ChatGPT-3) in oral and maxillofacial radiology for report writing by identifying radiographic anatomical landmarksand learning about oral and maxillofacial pathologiesand their radiographic features. The study also aimed to evaluate the performance of ChatGPT-3 and its usage in oral and maxillofacial radiology training. Materials and methods A questionnaire consisting of 80 questions was queried on the OpenAI app ChatGPT-3. The questions were stratified based on three categories. The categorization was based on random anatomical landmarks, oral and maxillofacial pathologies, and the radiographic features of some of these pathologies. One oral and maxillofacial radiologist evaluated queries that were answered by the ChatGPT-3 model and rated them on a 4-point, modified Likert scale. The post-survey analysis for the performance of ChatGPT-3 was based on the Strengths, Weaknesses, Opportunities, and Threats (SWOT) analysis,its application in oral and maxillofacial radiology training, and its recommended use. Results In order of efficiency, Chat GPT-3 gave 100%accuracy in describing radiographic landmarks. However, the content of the oral and maxillofacial pathologies was limited to major or characteristic radiographic features. The mean scores for the queries related to the anatomic landmarks, oral and maxillofacial pathologies, and radiographic features of the oral and maxillofacial pathologies were 3.94, 3.85, and 3.96, respectively. However, the median and mode scores were 4 and were similar to all categories. The data for the oral and maxillofacial pathologies when the questions were not specifically included in the format of the introduction of the pathology, causes, symptoms, and treatment. Out of two abbreviations, one was not answered correctly. Conclusion The study showed that ChatGPT-3 is efficient in describing the pathology, characteristic radiographic features, and describing anatomical landmarks. ChatGPT-3 can be used as an adjunct when an oral radiologist needs additional information on any pathology, however, it cannot be the mainstay for reference. ChatGPT-3 is less detail-oriented, and the data has a risk of infodemics and the possibility of medical errors. However, Chat GPT-3 can be an excellent tool in helping the community in increasing the knowledge and awareness of various pathologies and decreasing the anxiety of the patients while dental healthcare professionals formulate an appropriate treatment plan.

Poster 132: Femoral length on lateral knee radiographs influences accuracy of radiographic landmarks for MPFC reconstruction

Objectives: Accuracy of femoral tunnel positioning is critical during medial patellofemoral complex (MPFC) reconstruction, as even 5mm of malpositioning has been associated with altered patellofemoral contact pressures. To help identify the appropriate position for femoral tunnel placement, Schottle and others have described radiographic landmarks to identify the MPFC footprint on lateral knee radiographs. These measurements are based on an extension of the posterior cortical line, which can vary based on the length of the femur visible on the radiograph. Because the scope of view can vary between intraoperative imaging modalities, the aim of this study was to assess the effect of femoral length on the accuracy of radiographic landmarks of the MPFC. Methods: In 9 unpaired cadaveric knees, the MPFC footprint was exposed on the medial femur, and the proximal and distal boundaries of the footprint were marked. Lateral fluoroscopic images of the knee were obtained and assessed in 1 cm length increments, beginning 1cm proximal to the posterior condyle and continuing proximally to a femoral length of 8 cms. The MPFC midpoint was described on each image relative to the posterior cortical line of the femur and a line perpendicular to this line, relative to the proximal margin of the posterior condyle. Linear regression analysis was used to assess the effect of femoral length on the radiographic position of the MPFC. ROC curve analysis and Delong test were used, and the minimum amount of femoral length required on radiographs to accurately identify an anatomic femoral tunnel was determined using Youden’s J statistic. Results: Using the posterior cortical line as a reference, the radiographic description of the MPFC footprint moved anteriorly with decreasing femoral length on the radiographs, particularly at 4cm and less. However, no proximal-distal change was seen in relation to the line through the proximal margin of the posterior condyle with changing femoral lengths. Linear regression analysis showed a significant relationship between the femoral length and anterior radiographic position of the MPFC on radiographs (R = 0.461, R2 = 0.212, B= -0.636, p &lt; 0.001). The slope coefficient was -0.636 mm, indicating that for every cm decrease in femoral length, the actual anatomic footprint of the MPFC moves anteriorly by 0.636 mm in relation to the posterior cortical line. Furthermore, ROC curve analysis revealed that a minimum of 4 cm of femoral length on lateral radiographs is required to accurately localize the footprint of the MPFC (AUC 0.80; sensitivity 76.7 %; specificity 69 %; p &lt; 0.001). Conclusions: The radiographic landmarks for the MPFC femoral footprint can change depending on the length of the distal femur visible on radiographs. We found that at least 4cm of the femoral shaft should be visible for the radiographic landmarks to be accurate. As fluoroscopy is frequently used intraoperatively for MPFC reconstruction, our findings may serve as a guide for accurate femoral tunnel placement.

The Use of the Piriformis Fossa Radiographic Landmark to Predict "In-Out-In" Placement of the Posterosuperior Femoral Neck Screw.

To investigate the correlation between a screw's radiographic relationship to the piriformis fossa with position on CT in the clinical setting. Intraoperative fluoroscopic images of patients treated with cannulated screw fixation of a femoral neck fracture, who also had a post-operative CT scan, were retrospectively evaluated by four fellowship trained orthopedic trauma surgeons. The posterosuperior screw on the AP fluoroscopic view was determined to be either above the piriformis fossa (APF) or below the piriformis fossa (BPF). Using CT scan to determine IOI placement, the ability to predict IOI position based on fluoroscopic imaging was evaluated by calculating accuracy, sensitivity, specificity, and interobserver reliability. 73 patients met inclusion criteria. The incidence of IOI screw placement was 59% on CT evaluation. The use of the PF landmark accurately predicted CT findings in 89% of patients. A screw placed APF was 90% sensitive and 88% specific in predicting cortical breach, with near perfect interobserver agreement (κ = 0.81). The use of the PF radiographic landmark is highly sensitive and specific in predicting the placement of an IOI posterosuperior femoral neck screw.

Poster 118: False Profile Radiograph Sourcil-Edge and Bone-Edge Measurements Both Correlate to Different Weight-Bearing Regions of the Acetabulum: A Three-Dimensional Analysis

Objectives: The acetabular sourcil has been thought of as the radiographic representation of the weightbearing dome of the acetabulum; however, there are limited modern data to support this supposition. In the literature, assessment of weightbearing acetabular coverage has been described using both the sourcil-edge and bone-edge as anatomic landmarks, leading to confusion and controversy in classification of hip dysplasia, indications for hip preservation surgery, and intraoperative coverage assessment. The purpose of this study was to localize the three-dimensional (3D) anatomic correlates of the sourcil-edge and bone-edge radiographic position on false profile radiographs. The authors hypothesized that the sourcil-edge on the false profile represents anterolateral coverage while the bone-edge represents anterior coverage. Methods: Forty hips were grouped by large or small differences between bone-edge and sourcil-edge anterior center edge angles (ACEA). The small difference group comprised the lowest quartile of sourcil-edge to bone-edge difference of an initial cohort of 80 hips, while the large difference group comprised the highest quartile of difference. 3D surface mesh models and digitally reconstructed radiographs (DRRs) were generated from hip computed tomography scans using the 3D Slicer Image Computing Platform (Figures 1 &amp; 3). Sourcil-edge, bone-edge, and ACEAs were identified on DRRs and registered to the 3D models with fiducial markers. Intersection of bone-edge and sourcil-edge projection lines with the acetabular rim were obtained from the 3D models (Figures 2 &amp; 4). 3D Anterior acetabular rim morphology was correlated with ACEA difference. Results: The bone-edge and sourcil-edge projections intersected the acetabular rim at a mean of 2:08 ± 0:25 and 1:20 ± 0:24 o’clock, respectively. The 3D models consistently demonstrated that, in both large and small discrepancy groups, the sourcil-edge corresponded to an area just posterior to the AIIS projection while the bone-edge corresponded to the weight-bearing region inferior to the AIIS (Figures 2 &amp; 4). Additionally, in large discrepancy hips, the bone-edge corresponded to a more prominent acetabular coverage in the region of the AIIS than when compared to the small discrepancy hips. Conclusions: This study demonstrates that both the sourcil-edge and bone-edge projections on false profile radiographs correlate with differing weightbearing regions of the acetabulum. The sourcil-edge ACEA represents coverage at 1:20 o’clock on the acetabular rim and is consistently posterior to the AIIS, while the bone-edge measurement represents the region of the acetabular rim at 2:08 o’clock and is consistently inferior to the AIIS. In hips with a large discrepancy between bone-edge and sourcil-edge (highest quartile), ACEA measurement differs by a mean of 19.7 degrees based on the selection of these radiographic landmarks. These findings have substantial implications for hip preservation surgery technique and clinical decision making. A comprehensive understanding of femoral head coverage is essential for the treatment of young adult hip pathology. Although the sourcil and bone edge are two of the most important radiographic landmarks for assessment of acetabular dysplasia, descriptions of these landmarks are often over simplified. Some authors have advocated for greater utilization of measurements of coverage to the bone edge as it can be more consistently identified. Other authors have advocated for utilizing the sourcil-edge measurement as this is more representative of the “weight bearing dome”. These descriptions are limited as they do not improve our understanding of the anatomic and radiographic structure of the anterior acetabular rim. Although the concept of the “weight bearing dome” is classically described, it lacks anatomic precision. The present study clarifies the anatomic features of the acetabular rim that correspond to these radiographic features. Increased difference between the sourcil-edge and bone edge is generated by the differences in undulation of the anterosuperior acetabular rim. Acetabuli that have maximal anterior rim prominence at or near the superior-most aspect of the acetabulum (generally counterclockwise from the AIIS) tend to demonstrate smaller discrepancy between the bone-edge and sourcil-edge (Figure 4). Acetabuli that have maximal rim prominence medial and anterior to the superior-most aspect of the acetabulum generally demonstrate larger discrepancy (Figure 2). The present anatomic description can be utilized by the hip arthroscopist to better determine which areas of the anterior rim may cause impingement. Similarly, while conducting a periacetabular osteotomy, the surgeon may use this information to better understand how acetabular rotation is contributing to both anterior and anterolateral coverage. In conclusion, on false profile radiographs, the sourcil-edge corresponds to superior femoral head coverage, while the bone-edge corresponds to anterosuperior coverage. False profile radiographs with a large discrepancy between sourcil-edge and bone-edge measurements demonstrate a prominent weight bearing surface in the region of the AIIS. These data regarding the radiographic phenomena of the acetabular rim can aid in clinical decision-making and intraoperative technique.

Poster 129: Biomechanical Evaluation of the Isometry of Medial Patellofemoral Ligament Reconstruction Grafts Fixed at Non-Anatomic Femoral Insertion Points

Objectives: In MPFL reconstruction surgery, achieving perfect placement of the femoral tunnel intraoperatively is difficult. An understanding of the effects femoral tunnel malpositioning has on graft isometry is crucial to minimizing instability or overconstraint which leads to anterior knee pain, increased patellofemoral contact pressures, or graft failure. The purpose of this study was to identify differences in MPFL reconstruction graft isometry with femoral tunnel malpositioning, specifically evaluating isometry differences as the femoral position is moved anterior, posterior, proximal, and distal relative to the femoral radiographic landmark of the MPFL, Schottle’s point. Methods: A biomechanical study evaluating 11 fresh frozen cadaveric knees was conducted. Non-elastic suture, used as an MPFL graft analogue, was anchored at Schottle’s point with the knee at 30° of flexion, and at 5 mm and 10 mm anterior, posterior, superior, and distal to Schottle’s point. A draw wire displacement sensor was used to evaluate length changes of the MPFL graft analogue through 0-120° of knee flexion. Knee flexion position was continuously measured using a motion tracking system. Pairwise t-tests with Bonferroni corrections were used to compare isometry between Schottle’s point and non- anatomic femoral insertion points. Results: Grafts placed at Schottle’s point demonstrated mild anisometry with tightening in extension and loosening in flexion. Similarly, grafts placed distally and posteriorly also demonstrated tightening in extension and loosening in flexion. Grafts placed anteriorly and proximally demonstrated tightening in flexion. Mean and range of total draw wire excursion over full ROM across specimens is shown in Figure 1. The mean draw wire length was found to be greater with further displacement from Schottle’s point in all directions but posterior. Pairwise comparisons relative to Schottle’s point found grafts placed in the proximal or distal direction demonstrated significant differences in MPFL excursion, while grafts placed in the anterior or posterior direction did not (10 mm proximal: 0.36 (p=0.028), 5 mm distal: 0.14 (p=0.012), 10 mm distal: 0.22 (p&lt;0.001)) (Table 1). Conclusions: In order to minimize the risk of graft tightening during knee flexion, if deviating from Schottle’s point, it is safest to err posterior or distal in femoral tunnel position. Travel away from Schottle’s point in the proximal or distal direction has the greatest effects on graft isometry. The greatest risk of overconstraint and graft tightening in flexion is seen with proximal placement of the femoral insertion point.

A Comperative Study of Use Of Artificial Intelligence in Oral Radiology Education

Purpose: The aim of this study is to compare the efficacy of artificial intelligence use in oral radiology learning in the undergraduate dental students. Materials&amp;amp;Methods: Fifty third-year students in the University of Lokman Hekim were detected images with the artificial intelligence method (AI) and standard lecture method (SL) for anatomical landmarks in panoramic radiographs. SL consisted of a frontal lecture through a standardized presentation. CranioCatch model (Eskisehir, Turkey) was used as deep learning-based artificial intelligence model. One panoramic image was loaded to the application and anatomic landmarks were detected by teacher, students were asked to mark. AI recorded and scored students answers. A questionnaire study was conducted for the perception of students in terms of validity and reliability regarding assessment and evaluation for each methods. Results: 50 undergraduate students (26 female,24 male) answered 7questions, 5-point Likert type. The conformity to the normal distribution was evaluated with the Shapiro-Wilk test and the graphical approach (Normal Q-Q Plot). The values did not conform to the normal distribution. As a result of the reliability analysis performed for the measurement tool, the Cronbach’s Alpha coefficient was found 0.828. Wilcoxon Test was used to test the significance of the difference between each methods. There is a statistically significant difference between the mean values of evaluation measurements(p=0.014). AI was higher than the mean of evaluation measurement values compared to SL. Conclusion: AI models have performed very well in measurement and evaluation in oral radiology learning.

Evaluation of Position of Mental Foramen in Dentulous Population by Digital Panoramic Radiography.

The mental branch of the trigeminal nerve, which supplies sensation to the lower lip, originates in the mandibular canal, making it an essential anatomical structure for dentists and oral surgeons to access. It is not well known that there is a spectrum of normal that includes variants in which there may be more than one nerve entry site, which means that if the mental foramen is not protected, normal feeling in the lower lip may be lost permanently. The diagnostic value of global radiographic landmarks like the mandibular trench and the dental foramen in identifying skeletal problems has been investigated. Four hundred patients over 40 who fit the inclusion/exclusion criteria were chosen. Patients were recruited from Vananchal Dental College and Hospital, Garhwa's Out Patient section for Oral Medicine, Diagnosis, and Radiology. The following patients with their consent are subjected for digital orthopantomography (Cephalometric Device for Rotograph EVOD Ref 930790001, SN 14112930, Villa Sistem Medical) and the captured images are then interpreted for the variations in mandibular canal. The results of our study are as follows: In this study, the mean age of the subjects was 47.27 years, with a range of 40-75 yearsMajority of cases were females (53.75%) and 46.25% were males.The top of the residual ridge, located between the mental foramen and the mandibular canal, is clearly visible to all observers (grade I).Only 1.75% of the people surveyed had a bifid canal, despite the fact that the majority of the people surveyed had a single mandibular canal on both sides (98.25%).Right side mean was 18,682.017, whereas left side mean was 16,331.851; nevertheless, this difference was not statistically significant (P = 0.0860 NS). The dental foramen was located close to where the mandible and the area housing the next premolar met. These findings may be utilized to improve the safety of peri-apical surgical procedures. Therefore, it is therapeutically relevant to get insight into the structural alterations of the mental foramen and locate its location in preoperative radiological scans.

Paper 06: Femoral length on lateral knee radiographs influences accuracy of radiographic landmarks for MPFC reconstruction

Objectives: Accuracy of femoral tunnel positioning is critical during medial patellofemoral complex (MPFC) reconstruction, as even 5mm of malpositioning has been associated with altered patellofemoral contact pressures. To help identify the appropriate position for femoral tunnel placement, Schottle and others have described radiographic landmarks to identify the MPFC footprint on lateral knee radiographs. These measurements are based on an extension of the posterior cortical line, which can vary based on the length of the femur visible on the radiograph. Because the scope of view can vary between intraoperative imaging modalities, the aim of this study was to assess the effect of femoral length on the accuracy of radiographic landmarks of the MPFC. Methods: In 9 unpaired cadaveric knees, the MPFC footprint was exposed on the medial femur, and the proximal and distal boundaries of the footprint were marked. Lateral fluoroscopic images of the knee were obtained and assessed in 1 cm length increments, beginning 1cm proximal to the posterior condyle and continuing proximally to a femoral length of 8 cms. The MPFC midpoint was described on each image relative to the posterior cortical line of the femur and a line perpendicular to this line, relative to the proximal margin of the posterior condyle. Linear regression analysis was used to assess the effect of femoral length on the radiographic position of the MPFC. ROC curve analysis and Delong test were used to determine ROC curve analysis and Delong test were used, and the minimum amount of femoral length required on radiographs to accurately identify an anatomic femoral tunnel was determined using Youden’s J statistic. Results: Using the posterior cortical line as a reference, the radiographic description of the MPFC footprint moved anteriorly with decreasing femoral length on the radiographs, particularly at 4cm and less. However, no proximal-distal change was seen in relation to the line through the proximal margin of the posterior condyle with changing femoral lengths. Linear regression analysis showed a significant relationship between the femoral length and anterior position of the MPFC on radiographs (R = 0.461, R2 = 0.212, B= -0.636, p &lt; 0.001). The slope coefficient was -0.636 mm, indicating that for every cm decrease in femoral length, the actual anatomic footprint of the MPFC moves anteriorly by 0.636 mm in relation to the posterior cortical line. Furthermore, ROC curve analysis revealed that a minimum of 4 cm of femoral length on lateral radiographs is required to accurately localize the footprint of the MPFC (AUC 0.79; sensitivity 76.7 %; specificity 69 %; p &lt; 0.001). Conclusions: The radiographic landmarks for the MPFC femoral footprint can change depending on the length of the distal femur visible on radiographs. We found that at least 4cm of the femoral shaft should be visible for the radiographic landmarks to be accurate. As fluoroscopy is frequently used intraoperatively for MPFC reconstruction, our findings may serve as a guide for accurate femoral tunnel placement.

Automated diagnosis of flatfoot using cascaded convolutional neural network for angle measurements in weight-bearing lateral radiographs.

Diagnosis of flatfoot using a radiograph is subject to intra- and inter-observer variabilities. Here, we developed a cascade convolutional neural network (CNN)-based deep learning model (DLM) for an automated angle measurement for flatfoot diagnosis using landmark detection. We used 1200 weight-bearing lateral foot radiographs from young adult Korean males for the model development. An experienced orthopedic surgeon identified 22 radiographic landmarks and measured three angles for flatfoot diagnosis that served as the ground truth (GT). Another orthopedic surgeon (OS) and a general physician (GP) independently identified the landmarks of the test dataset and measured the angles using the same method. External validation was performed using 100 and 17 radiographs acquired from a tertiary referral center and a public database, respectively. The DLM showed smaller absolute average errors from the GT for the three angle measurements for flatfoot diagnosis compared with both human observers. Under the guidance of the DLM, the average errors of observers OS and GP decreased from 2.35° ± 3.01° to 1.55° ± 2.09° and from 1.99° ± 2.76° to 1.56° ± 2.19°, respectively (both p < 0.001). The total measurement time decreased from 195 to 135min in observer OS and from 205 to 155min in observer GP. The absolute average errors of the DLM in the external validation sets were similar or superior to those of human observers in the original test dataset. Our CNN model had significantly better accuracy and reliability than human observers in diagnosing flatfoot, and notably improved the accuracy and reliability of human observers. • Development of deep learning model (DLM) that allows automated angle measurements for landmark detection based on 1200 weight-bearing lateral radiographs for diagnosing flatfoot. • Our DLM showed smaller absolute average errors for flatfoot diagnosis compared with two human observers. • Under the guidance of the model, the average errors of two human observers decreased and total measurement time also decreased from 195 to 135min and from 205 to 155min.

CORRELATION ANALYSIS OF DIFFERENT PELVIC TILT DEFINITIONS: A PRELIMINARY STUDY

Pelvic tilt (PT) is always described as the pelvic orientation along the transverse axis, yet four PT definitions were established based on different radiographic landmarks: anterior pelvic plane (PTa), the centres of femoral heads and sacral plate (PTm), pelvic outlet (PTh), and sacral slope (SS). These landmarks quantify a similar concept, yet understanding of their relationships is lacking. Some studies referred to the words “pelvic tilt” for horizontal comparisons, but their PT definitions might differ. There is a demand for understanding their correlations and differences for education and research purposes.This study recruited 105 sagittal pelvic radiographs (68 males and 37 females) from a single clinic awaiting their hip surgeries. Hip hardware and spine pathologies were examined for sub-group analysis. Two observers annotated four PTs in a gender-dependent manner and repeated it after six months. The linear regression model and intraclass correlation coefficient (ICC) were applied with a 95% significance interval.The SS showed significant gender differences and the lowest correlations to the other parameters in the male group (−0.3&lt; r &lt;0.2). The correlations of SS in scoliosis (n = 7) and hip implant (female, n = 18) groups were statistically different, yet the sample sizes were too small. PTm demonstrated very strong correlation to PTh (r &gt; 0.9) under the linear model PTm = 0.951 × PTh - 68.284.The PTm and PTh are interchangeable under a simple linear regression model, which enables study comparisons between them. In the male group, SS is more of a personalised spinal landmark independent of the pelvic anatomy. Female patients with hip implant may have more static spinopelvic relationships following a certain pattern, yet a deeper study using a larger dataset is required. The understanding of different PTs improves anatomical education.