Pelvic incidence and pelvic tilt measurements using femoral heads or acetabular domes to identify centers of the hips: comparison of two methods

Automated measurement of pelvic parameters using convolutional neural network in complex spinal deformities: overcoming challenges in coronal deformity cases

BACKGROUND CONTEXT

Accurate and consistent measurement of sagittal alignment is challenging, particularly in patients with severe coronal deformities, including degenerative lumbar scoliosis (DLS).

PURPOSE

This study aimed to develop and validate an artificial intelligence (AI)-based system for automating the measurement of key sagittal parameters, including lumbar lordosis, pelvic incidence, pelvic tilt, and sacral slope, with a focus on its applicability across a wide range of deformities, including severe coronal deformities, such as DLS.

DESIGN

Retrospective observational study.

PATIENT SAMPLE

A total of 1,011 standing lumbar lateral radiographs, including DLS.

OUTCOME MEASURE

Interclass and intraclass correlation coefficients (CC), and Bland-Altman plots.

METHODS

The model utilizes a deep-learning framework, incorporating a U-Net for segmentation and a Keypoint Region-based Convolutional Neural Network for keypoint detection. The ground truth masks were annotated by an experienced orthopedic specialist. The performance of the model was evaluated against ground truth measurements and assessments from two expert raters using interclass and intraclass CC, and Bland-Altman plots.

RESULTS

In the test set of 113 patients, 39 (34.5%) had DLS, with a mean Cobb's angle of 14.8°±4.4°. The AI model achieved an intraclass CC of 1.00 across all parameters, indicating perfect consistency. Interclass CCs comparing the AI model to ground truth ranged from 0.96 to 0.99, outperforming experienced orthopedic surgeons. Bland-Altman analysis revealed no significant systemic bias, with most differences falling within clinically acceptable ranges. A 5-fold cross-validation further demonstrated robust performance, with interclass CCs ranging from 0.96 to 0.99 across diverse subsets.

CONCLUSION

This AI-based system offers a reliable and efficient automated measurement of sagittal parameters in spinal deformities, including severe coronal deformities. The superior performance of the model compared with that of expert raters highlights its potential for clinical applications.

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.

Inadequate Annotation and Its Impact on Pelvic Tilt Measurement in Clinical Practice

BACKGROUND

Accurate pre-surgical templating of the pelvic tilt (PT) angle is essential for hip and spine surgeries, yet the reliability of PT annotations is often compromised by human error, inherent subjectivity, and variations in radiographic quality. This study aims to identify challenges leading to inadequate annotations at a landmark dimension and evaluating their impact on PT.

METHODS

We retrospectively collected 115 consecutive sagittal radiographs for the measurement of PT based on two definitions: the anterior pelvic plane and a line connecting the femoral head's centre to the sacral plate's midpoint. Five annotators engaged in the measurement, followed by a secondary review to assess the adequacy of the annotations across all the annotators.

RESULTS

The outcomes indicated that over 60% images had at least one landmark considered inadequate by the majority of the reviewers, with poor image quality, outliers, and unrecognized anomalies being the primary causes. Such inadequacies led to discrepancies in the PT measurements, ranging from -2° to 2°.

CONCLUSION

This study highlights that landmarks annotated from clear anatomical references were more reliable than those estimated. It also underscores the prevalence of suboptimal annotations in PT measurements, which extends beyond the scope of traditional statistical analysis and could result in significant deviations in individual cases, potentially impacting clinical outcomes.

Concurrent Learning Approach for Estimation of Pelvic Tilt from Anterior-Posterior Radiograph

Accurate and reliable estimation of the pelvic tilt is one of the essential pre-planning factors for total hip arthroplasty to prevent common post-operative complications such as implant impingement and dislocation. Inspired by the latest advances in deep learning-based systems, our focus in this paper has been to present an innovative and accurate method for estimating the functional pelvic tilt (PT) from a standing anterior-posterior (AP) radiography image. We introduce an encoder-decoder-style network based on a concurrent learning approach called VGG-UNET (VGG embedded in U-NET), where a deep fully convolutional network known as VGG is embedded at the encoder part of an image segmentation network, i.e., U-NET. In the bottleneck of the VGG-UNET, in addition to the decoder path, we use another path utilizing light-weight convolutional and fully connected layers to combine all extracted feature maps from the final convolution layer of VGG and thus regress PT. In the test phase, we exclude the decoder path and consider only a single target task i.e., PT estimation. The absolute errors obtained using VGG-UNET, VGG, and Mask R-CNN are 3.04 ± 2.49, 3.92 ± 2.92, and 4.97 ± 3.87, respectively. It is observed that the VGG-UNET leads to a more accurate prediction with a lower standard deviation (STD). Our experimental results demonstrate that the proposed multi-task network leads to a significantly improved performance compared to the best-reported results based on cascaded networks.

Repeatability of Inertial Measurement Units for Measuring Pelvic Mobility in Patients Undergoing Total Hip Arthroplasty

Consideration of pelvic mobility when positioning implants for total hip arthroplasty (THA) has been shown to reduce the risk of complications such as dislocation, squeaking and excessive wear. We aim to test the repeatability of pelvic tilt measurements taken between three positions (standing, flexed-seated and step-up) by an inertial measurement unit (IMU) and hence, evaluate their reliability in screening for high pelvic mobility in patients undergoing THA. The repeated IMU measurements of pelvic tilt were analysed for consistency and compared with measures taken by x-ray analysis. Our study showed greater variation in measures taken by the IMU particularly in the flexed-seated position. The patient's pelvic tilt in this position negatively correlated with their mid-back angle, suggesting the posture of the patient is a source of variation in the flexed-seated position if not kept consistent during assessments. IMUs were overall able to produce accurate and reliable measurements of pelvic tilt; however, protocols will need to be adjusted to factor in a patient's mid-back angle when taking future readings.

Spinopelvic Parameters Do Not Influence Outcomes Following Primary Hip Arthroscopy for the Treatment of Femoroacetabular Impingement Syndrome

Purpose

To evaluate the influence of spinopelvic parameters on short-term postoperative patient-reported outcomes (PROs) following primary hip arthroscopy for the treatment of femoroacetabular impingement syndrome (FAIS).

Methods

Patients undergoing primary hip arthroscopy between January 2012 and December 2015 were retrospectively reviewed. Hip Outcome Score - Activities of Daily Living, Hip Outcome Score - Sports-Specific Subscale, modified Harris Hip Score, International Hip Outcome Tool-12, and visual analog scale pain were recorded preoperatively and at final follow-up. Lumbar lordosis (LL), pelvic tilt (PT), sacral slope, and pelvic incidence (PI) were measured on lateral radiographs in standing position. Patients were split into subgroups for individual analyses based on previous literature cutoffs: |PI-LL|> or <10°, PT> or <20°, and PI <40°, 40° < PI < 65°, and PI >65°. PROs and rate of achievement of patient acceptable symptom state (PASS) were compared between subgroups at final follow-up.

Results

Sixty-one patients who underwent unilateral hip arthroscopy were included in the analysis, and 66% of patients were female. Mean patient age was 37.6 ± 11.3 years, whereas mean body mass index was 25.0 ± 5.7. Mean follow-up time was 27.6 ± 9.0 months. No significant difference in preoperative nor postoperative PROs were appreciated in patients with spinopelvic mismatch (|PI-LL| >10°) versus those without, whereas patients with mismatch achieved PASS according to the modified Harris Hip Score (P = .037) and International Hip Outcome Tool-12 (P = .030) at greater rates. When we compared patients with a PT ≥20° versus PT <20°, no significant differences in postoperative PROs were present. When we compared patients in the following pelvic incidence groups: PI <40°, 40° < PI <65°, and PI >65°, no significant differences in 2-year PROs or rates of PASS achievement for any PRO were appreciated (P > .05 for all).

Conclusions

In this study, spinopelvic parameters and traditional measures of sagittal imbalance did not influence PROs in patients undergoing primary hip arthroscopy for FAIS. Patients with sagittal imbalance (|PI-LL|> 10° or PT >20°) achieved a greater rate of PASS.

Level of Evidence

IV; Prognostic case series.

Artificial Intelligence System for Automatic Quantitative Analysis and Radiology Reporting of Leg Length Radiographs

Leg length discrepancies are common orthopedic problems with the potential for poor functional outcomes. These are frequently assessed using bilateral leg length radiographs. The objective was to determine whether an artificial intelligence (AI)-based image analysis system can accurately interpret long leg length radiographic images. We built an end-to-end system to analyze leg length radiographs and generate reports like radiologists, which involves measurement of lengths (femur, tibia, entire leg) and angles (mechanical axis and pelvic tilt), describes presence and location of orthopedic hardware, and reports laterality discrepancies. After IRB approval, a dataset of 1,726 extremities (863 images) from consecutive examinations at a tertiary referral center was retrospectively acquired and partitioned into train/validation and test sets. The training set was annotated and used to train a fasterRCNN-ResNet101 object detection convolutional neural network. A second-stage classifier using a EfficientNet-D0 model was trained to recognize the presence or absence of hardware within extracted joint image patches. The system was deployed in a custom web application that generated a preliminary radiology report. Performance of the system was evaluated using a holdout 220 image test set, annotated by 3 musculoskeletal fellowship trained radiologists. At the object detection level, the system demonstrated a recall of 0.98 and precision of 0.96 in detecting anatomic landmarks. Correlation coefficients between radiologist and AI-generated measurements for femur, tibia, and whole-leg lengths were > 0.99, with mean error of < 1%. Correlation coefficients for mechanical axis angle and pelvic tilt were 0.98 and 0.86, respectively, with mean absolute error of < 1°. AI hardware detection demonstrated an accuracy of 99.8%. Automatic quantitative and qualitative analysis of leg length radiographs using deep learning is feasible and holds potential in improving radiologist workflow.

Can Hip-Knee Line Angle Distinguish the Size of Pelvic Incidence?-Development of Quick Noninvasive Assessment Tool for Pelvic Incidence Classification

This study aimed to explore effective measurement angles for pelvic incidence (PI) classification and to develop a quick, noninvasive assessment tool for PI classification. We defined five variation types of hip–knee line (HKL) angles and tested the discrimination ability of the receiver operating characteristic (ROC) analysis using 125 photographs of upright standing posture from the right lateral side. ROC analysis revealed an applicable HKL angle defined by the line connecting the most raised part of the buttock and the central point of the knee and the midthigh line. The acceptable cut-off points for discriminating small or large PIs in terms of HKL angle were 18.5° for small PI (sensitivity, 0.91; specificity, 0.79) and 21.5° for large PI discrimination (sensitivity, 0.74; specificity, 0.72). In addition, we devised a quick noninvasive assessment tool for PI classification using the cut-offs of the HKL angle with a view to practical application. The results of intra- and inter-rater reliability ensured a substantial/moderate level of the tool (Cohen’s kappa coefficient, 0.79; Fleiss’s kappa coefficient, 0.50–0.54). These results revealed that the HKL angle can distinguish the size of the PI with a high/moderate discrimination ability. Furthermore, the tool indicated acceptable inter-/intra-rater reliability for practical applications.

Pelvic incidence and pelvic tilt can be calculated using either the femoral heads or acetabular domes in patients with hip osteoarthritis

AIMS

The aim of the study was to compare two methods of calculating pelvic incidence (PI) and pelvic tilt (PT), either by using the femoral heads or acetabular domes to determine the bicoxofemoral axis, in patients with unilateral or bilateral primary hip osteoarthritis (OA).

METHODS

PI and PT were measured on standing lateral radiographs of the spine in two groups: 50 patients with unilateral (Group I) and 50 patients with bilateral hip OA (Group II), using the femoral heads or acetabular domes to define the bicoxofemoral axis. Agreement between the methods was determined by intraclass correlation coefficient (ICC) and the standard error of measurement (SEm). The intraobserver reproducibility and interobserver reliability of the two methods were analyzed on 31 radiographs in both groups to calculate ICC and SEm.

RESULTS

In both groups, excellent agreement between the two methods was obtained, with ICC of 0.99 and SEm 0.3° for Group I, and ICC 0.99 and SEm 0.4° for Group II. The intraobserver reproducibility was excellent for both methods in both groups, with an ICC of at least 0.97 and SEm not exceeding 0.8°. The study also revealed excellent interobserver reliability for both methods in both groups, with ICC 0.99 and SEm 0.5° or less.

CONCLUSION

Either the femoral heads or acetabular domes can be used to define the bicoxofemoral axis on the lateral standing radiographs of the spine for measuring PI and PT in patients with idiopathic unilateral or bilateral hip OA. Cite this article: Bone Joint J 2021;103-B(8):1345-1350.

Assessment of sagittal spinopelvic parameters in a Taiwanese population with spondylolysis by the EOS imaging system: a retrospective radiological analysis

Background

The impact of sagittal spinopelvic alignment on spondylolysis is well established in Caucasian populations. However, prior studies suggest that people from different ethnological backgrounds showed divergence, and a few studies that focused on Asian populations reported conflicting results. The aim of this study is to use the EOS imaging system to evaluate the spinopelvic parameters of spondylolysis patients, and their relationship with spondylolisthesis, disc degeneration, and age in a Taiwanese population.

Methods

Radiographic sagittal spinopelvic parameters for 45 spondylolysis patients and 32 healthy people were evaluated, including pelvic incidence (PI), sacral slope (SS), pelvic tilt (PT), thoracic kyphosis (TK), and lumbar lordosis (LL). The spinopelvic parameters were compared between spondylolytic and control groups. These parameters were further compared between spondylolytic subjects with and without spondylolisthesis, with and without high-grade disc degeneration, and young (< 30 years old) and middle-aged.

Results

The PI and LL of the spondylolytic group (52.6°±12.0° and 41.3°±15.2°) were significantly higher than those of the healthy control group (47.16°±7.95° and 28.22°±10.65°). Further analysis of the spondylolytic patients revealed that those with high-grade disc degeneration were more prone to spondylolisthesis (92.3 %) compared to those without (50 %; p = 0.001). The middle-aged group had significantly higher rates of spondylolisthesis (80 %) and high-grade disc degeneration (52.4 %) compared with those for the young group (45 and 16.7 %, respectively; p = 0.017 and 0.047, respectively). No statistically significant difference in the sagittal spinopelvic parameters was found when spondylolytic patients were divided according to the occurrence of spondylolisthesis or high-grade disc degeneration.

Conclusions

In a Taiwanese population, PI and LL were significantly larger in spondylolytic patients. Disc degeneration and age were associated with the occurrence of spondylolisthesis. Ethnological differences should thus be taken into account when making clinical decisions regarding spondylolysis in a Taiwanese population.

A Sensor-Based Screening Tool for Identifying High Pelvic Mobility in Patients Due to Undergo Total Hip Arthroplasty

There is increasing evidence that pelvic mobility is a critical factor to consider in implant alignment during total hip arthroplasty (THA). Here, we test the feasibility of using an inertial sensor fitted across the sacrum to measure change in pelvic tilt, and hence screen for patients with high pelvic mobility. Patients (n = 32, mean age: 57.4 years) due to receive THA surgery participated in the study. Measures of pelvic tilt were captured simultaneously using the device and radiograph in three functional positions: Standing, flexed-seated, and step-up. We found a strong correlation between the device and radiograph measures for the change in pelvic tilt measure from standing to flexed-seated position (R² = 0.911); 75% of absolute errors were under 5 degrees. We demonstrated that the device can be used as a screening tool to rapidly identify patients who would benefit from more detailed surgical planning of implant positioning to reduce future risks of impingement and dislocation.

2016 SOSORT guidelines: orthopaedic and rehabilitation treatment of idiopathic scoliosis during growth

BACKGROUND

The International Scientific Society on Scoliosis Orthopaedic and Rehabilitation Treatment (SOSORT) produced its first guidelines in 2005 and renewed them in 2011. Recently published high-quality clinical trials on the effect of conservative treatment approaches (braces and exercises) for idiopathic scoliosis prompted us to update the last guidelines' version. The objective was to align the guidelines with the new scientific evidence to assure faster knowledge transfer into clinical practice of conservative treatment for idiopathic scoliosis (CTIS).

METHODS

Physicians, researchers and allied health practitioners working in the area of CTIS were involved in the development of the 2016 guidelines. Multiple literature reviews reviewing the evidence on CTIS (assessment, bracing, physiotherapy, physiotherapeutic scoliosis-specific exercises (PSSE) and other CTIS) were conducted. Documents, recommendations and practical approach flow charts were developed using a Delphi procedure. The process was completed with the Consensus Session held during the first combined SOSORT/IRSSD Meeting held in Banff, Canada, in May 2016.

RESULTS

The contents of the new 2016 guidelines include the following: background on idiopathic scoliosis, description of CTIS approaches for various populations with flow-charts for clinical practice, as well as literature reviews and recommendations on assessment, bracing, PSSE and other CTIS. The present guidelines include a total of 68 recommendations divided into following topics: bracing (n = 25), PSSE to prevent scoliosis progression during growth (n = 12), PSSE during brace treatment and surgical therapy (n = 6), other conservative treatments (n = 2), respiratory function and exercises (n = 3), general sport activities (n = 6); and assessment (n = 14). According to the agreed strength and level of evidence rating scale, there were 2 recommendations on bracing and 1 recommendation on PSSE that reached level of recommendation "I" and level of evidence "II". Three recommendations reached strength of recommendation A based on the level of evidence I (2 for bracing and one for assessment); 39 recommendations reached strength of recommendation B (20 for bracing, 13 for PSSE, and 6 for assessment).The number of paper for each level of evidence for each treatment is shown in Table 8.

CONCLUSION

The 2016 SOSORT guidelines were developed based on the current evidence on CTIS. Over the last 5 years, high-quality evidence has started to emerge, particularly in the areas of efficacy of bracing (one large multicentre trial) and PSSE (three single-centre randomized controlled trials). Several grade A recommendations were presented. Despite the growing high-quality evidence, the heterogeneity of the study protocols limits generalizability of the recommendations. There is a need for standardization of research methods of conservative treatment effectiveness, as recognized by SOSORT and the Scoliosis Research Society (SRS) non-operative management Committee.

Influence of lower limbs discrepancy and pelvic coronal rotation on pelvic incidence, pelvic tilt and sacral slope

PURPOSE

Pelvic incidence (PI), pelvic tilt (PT) and sacral slope (SS) are important parameters in sagittal spine alignment evaluation. The measurements are a projection of the three-dimensional pelvis onto a two-dimensional radiograph and they may be influenced by orientation of the pelvis. The aim of this study was to assess the influence of pelvic rotation in the coronal plane (CPR) on radiographic accuracy of PI, PT, and SS measurements.

METHODS

Radiological evaluation of the CPR angel was performed on 1 radiological phantom. The radiographs were taken in 5° CPR increments over a range of 0°-45° (evaluated with a digital protractor). On each of the lateral radiograph, PI, PT, and SS were measured three times by three independent researchers. The lowest CPR that changed PI, PT, or SS by ≥6° (the highest reported error of measurement of these parameters) was considered as unacceptable. Next, CPR was calculated based on the distance between femoral heads (FHD). The agreement of the calculated and measured CPR was quantified by the intraclass correlation coefficient (ICC) and the median error for a single measurement (SEM), with value 0.75 considered as excellent agreement.

RESULTS

PI, PT and SS could be measured with an acceptable error of 6° on radiographs with up to 20° pelvic rotation. From 20° CPR onwards the S1 endplate was distorted, that makes the measurements of PI, PT and SS questionable. There was an excellent agreement between CPR measured with a protractor and calculated based on FHD with ICC of 0.99 and SEM of 1.1°.

CONCLUSIONS

Rotation of the pelvis in the coronal plane during acquisition of radiographs influences PI, PT and SS measurements. Substantial error of PI, PT and SS measurements occurs with CPR of more than 20° which is equivalent to a lower limb discrepancy of 5.2 cm. CPR may be calculated while acquiring the radiograph. Further evaluation of the influence of CPR on spinopelvic parameters with a larger sample would be valuable.

Role of T1 Pelvic Angle in Assessing Sagittal Balance in Outpatients With Unspecific Low Back Pain

The aim of the study was to explore the significance of T1 pelvic angle (TPA) for assessment of sagittal balance in a cohort of Chinese patients with unspecific low back pain. TPA has been commonly used to assess sagittal balance in adult spinal deformity. However, whether TPA could be used to assess sagittal balance in patients with unspecific low back pain effectively remains unanswered. Medical records of outpatients with unspecific low back pain who received treatment in our outpatient clinic between September 2013 and November 2014 were reviewed. Demographic data and radiographic data were collected. Correlation coefficients between TPA and other sagittal parameters were analyzed, and the intraclass correlation coefficient (ICC) analysis was performed to assess the inter- and intra-observer reliability of TPA. Patients were divided into 2 groups according to whether they were well-aligned (TPA ≤ 20°) or poorly aligned (TPA > 20°), and then demographic and sagittal parameters were compared between the 2 groups of patients. A total of 97 patients with unspecific low back pain were included in this study. The inter- and intraobserver reliability of the TPA measure had excellent agreement (ICC = 0.985 and 0.919, respectively). There were significant correlations between TPA and age, LL, PT, PI, T1SPI, SVA, and NRS (all P < 0.05). Of the 38 well-aligned patients in Group A, SVA was ≤5 cm in 33 (86.84%) patients and >5 cm in the other 5 (13.16%) patients, and of the 59 poorly aligned patients in Group B, SVA was >5 cm in 42 (71.19%) patients and ≤5 cm in the other 17 (28.81%) patients. There were significant differences in age, LL, SS, PT, PI, T1SPI, SVA, and NRS between the 2 groups of patients, but no significant difference was observed in TK and TL. TPA could be used to assess sagittal balance in outpatients with unspecific low back pain effectively.