Bone density measurements adjacent to acetabular cups in total hip arthroplasty using dual-energy CT: an in vivo reliability and agreement study

Change in CT-measured acetabular bone density following total hip arthroplasty: a systematic review and meta-analysis

BACKGROUND AND PURPOSE

Assessing peri-acetabular bone quality is valuable for optimizing the outcomes of primary total hip arthroplasty (THA) as preservation of good quality bone stock likely affects implant stability. The aim of this study was to perform a meta-analysis of peri-acetabular bone mineral density (BMD) changes over time measured using quantitative computer tomography (CT) and, second, to investigate the influence of age, sex, and fixation on the change in BMD over time.

METHODS

A systematic search of Embase, Scopus, Web of Science, and PubMed databases identified 19 studies that measured BMD using CT following THA. The regions of interest (ROI), reporting of BMD results, and scan protocols were extracted. A meta-analysis of BMD was performed on 12 studies that reported measurements immediately postoperatively and at follow-up.

RESULTS

The meta-analysis determined that periacetabular BMD around both cemented and uncemented components decreases over time. The amount of BMD loss increased relative to proximity of the acetabular component. There was a greater decrease in cortical BMD over time in females and cancellous BMD for young patients of any sex.

CONCLUSION

Peri-acetabular BMD decreases at different rates relative to its proximity to the acetabular component. Cancellous BMD decreases more in young patients and cortical bone decreases more in females. Standardized reporting parameters and suggested ROI to measure peri-acetabular BMD are proposed, to enable comparison between implant and patient variables in the future.

Prediction of bone mineral density in CT using deep learning with explainability

Bone mineral density (BMD) is a key feature in diagnosing bone diseases. Although computational tomography (CT) is a common imaging modality, it seldom provides bone mineral density information in a clinic owing to technical difficulties. Thus, a dual-energy X-ray absorptiometry (DXA) is required to measure bone mineral density at the expense of additional radiation exposure. In this study, a deep learning framework was developed to estimate the bone mineral density from an axial cut of the L1 bone on computational tomography. As a result, the correlation coefficient between bone mineral density estimates and dual-energy X-ray absorptiometry bone mineral density was .90. When the samples were categorized into abnormal and normal groups using a standard (T-score = - 1.0 ), the maximum F1 score in the diagnostic test was .875. In addition, it was identified using explainable artificial intelligence techniques that the network intensively sees a local area spanning tissues around the vertebral foramen. This method is well suited as an auxiliary tool in clinical practice and as an automatic screener for identifying latent patients in computational tomography databases.

A semiautomated method to quantitatively assess osteolytic lesion volume and bone mineral density within acetabular regions of interest from CT

The objectives of this study were to (1) develop a semiautomated method to obtain lesion volume and bone mineral density (BMD) in terms of Hounsfield units from pelvic computed tomography (CT) scans in three regions of interest, and (2) assess accuracy and reliability of the method based on cadaveric CT scans. Image artefacts due to metal implants reduce CT clarity and are more severe with more than one implant in situ. Therefore, accuracy and reliability tests were performed with varying numbers of total hip arthroplasties implanted. To test the accuracy of lesion size measurements, microcomputed tomography was used as a reference. Mean absolute error ranged from 36 to 284 mm³ after five measurements. Intra- and inter-operator reliability of the entire method was measured for a selection of parameters. All coefficient of variation values were good to excellent for CT scans of the native pelvic anatomy and a CT scans of the same pelvis with one and two implants in situ. Accuracy of quantifying lesion volume decreased with decreasing CT image clarity by 0.6%-3.6% mean absolute relative error. Reliability of lesion volume measurement decreased with decreasing CT clarity. This was also the case for reliability of BMD measurements in the region most disrupted by metal artefact. The presented method proposes an approach for quantifying bone loss which has been proven to be accurate, reliable, and clinically applicable.

Evaluation of peri-prosthetic radiolucent lines surrounding the cementless femoral stem using digital tomosynthesis with metal artifact reduction: a cadaveric study in comparison with radiography and computed tomography

Background

The traditional criterion for the diagnosis of implant loosening in total hip arthroplasty (THA) was once defined as a radiolucent line of >2 mm in width, based on plain radiography. Recent progress in imaging technology has allowed for the identification of complete radiolucent lines of ≤2 mm around the whole prosthesis as the basis for diagnosing component loosening in the absence of component migration. This study aimed to compare the sensitivity and specificity of digital tomosynthesis with metal artifact reduction with those of radiography and conventional computed tomography (CT) for detecting radiolucent lines of ≤2 mm surrounding cementless femoral stems of different widths.

Methods

The medullary canals of 4 cadaveric femurs were broached to 13 mm in diameter. Cylindrical cementless femoral stems with diameters of 9, 10, 11, 12, and 13 mm were sequentially inserted into each femur, creating 5 groups of radiolucent lines 2.0, 1.5-1.6, 1.1-1.2, 0.5-0.6, and 0 mm in diameter, respectively. Imaging by tomosynthesis, radiography, and CT was conducted for each radiolucent line model. The width information of the radiolucent line models was used as a reference standard for calculating sensitivity and specificity: observations in the group of 0 mm were used for calculating specificity, and those in the other four groups were used for sensitivity. The differences in sensitivity and specificity between the imaging methods were compared with chi-square test, and the 95% confidence intervals of improvements in the sensitivity and specificity of tomosynthesis compared with radiography and CT were calculated using mixed effects models.

Results

The overall sensitivity of tomosynthesis (63.3%) for detecting radiolucent lines ≤2 mm wide was significantly higher (P<0.017) than that of radiography (20.5%) and CT (50.2%), an improvement of 58.2%±3.1% (95% CI, P<0.001) and 21.7%±7.1% (95% CI, P<0.001) compared to radiography and CT, respectively. The sensitivity values for detecting radiolucent lines in all four groups by tomosynthesis and CT were significantly higher than those of radiography (P<0.017). Tomosynthesis also had significantly higher sensitivity than CT (P<0.017) in detecting radiolucent line ≤1.2 mm wide. The specificity of TMAR, radiography, and CT for detecting radiolucent lines was 87.5%, 92.5%, and 82.5%, respectively, with no significant difference (P>0.017).

Conclusions

Digital tomosynthesis with metal artifact reduction had significantly higher sensitivity than radiography for detecting radiolucent lines ≤2 mm wide surrounding cementless femoral stems. It also displayed higher sensitivity than CT for detecting radiolucent lines ≤1.2 mm in width. With a higher rate of detection for radiolucent lines narrower than 2 mm, tomosynthesis has the potential to improve the accuracy of early diagnosis of cementless THA stem loosening in clinical practice.

Dual-energy CT for automatic organs-at-risk segmentation in brain-tumor patients using a multi-atlas and deep-learning approach

In radiotherapy, computed tomography (CT) datasets are mostly used for radiation treatment planning to achieve a high-conformal tumor coverage while optimally sparing healthy tissue surrounding the tumor, referred to as organs-at-risk (OARs). Based on CT scan and/or magnetic resonance images, OARs have to be manually delineated by clinicians, which is one of the most time-consuming tasks in the clinical workflow. Recent multi-atlas (MA) or deep-learning (DL) based methods aim to improve the clinical routine by an automatic segmentation of OARs on a CT dataset. However, so far no studies investigated the performance of these MA or DL methods on dual-energy CT (DECT) datasets, which have been shown to improve the image quality compared to conventional 120 kVp single-energy CT. In this study, the performance of an in-house developed MA and a DL method (two-step three-dimensional U-net) was quantitatively and qualitatively evaluated on various DECT-derived pseudo-monoenergetic CT datasets ranging from 40 keV to 170 keV. At lower energies, the MA method resulted in more accurate OAR segmentations. Both the qualitative and quantitative metric analysis showed that the DL approach often performed better than the MA method.

Relative radiographic bone density measurement in revision hip arthroplasty and its correlation with qualitative subjective assessment by experienced surgeons

BACKGROUND

Conventional radiography (CR) is the imaging method of choice in monitoring bone remodelling and other stability parameters after total hip arthroplasty (THA). Quantitative roentgen- or computed-tomography-based methods to determine bone density are prone to metal artifacts and often very costly, which is why they are not used as standard in a clinical setting. Since subjective assessment of bone remodelling in CR also has a certain susceptibility to errors, semi-quantitative methods have been developed to help approximate periprosthetic bone density development via CR to open up an additional tool for documentation of radiographic THA follow-up.

OBJECTIVE

Proof-of-principle of a newly designed imaging-software-aided method to measure relative bone density around the femoral stem in a series of conventional radiographs following THA.

METHODS

Eighty-six patients with hip modular tapered, fluted titanium stems were selected from the clinical database and series of baseline and postoperative follow-up radiographs were obtained after 24 and 48 weeks. Relative bone densities were measured per Gruen zones G1-7 with the use of an open-source image analysis package (ImageJ) by means of greyscale histograms. In addition, subjective evaluation of selected cases was performed by three independent, blinded orthopedic surgeons. Besides descriptive and nonparametric analyses, intra-class correlation (ICC) was performed and objective and subjective results were compared by linear regression analysis.

RESULTS

Two individual cases are presented as a proof-of-principle. Increase or decrease of bone density could be measured correctly over time in each case. In a collective analysis there were no significant differences in mean relative bone densities between groups after 24 and 48 weeks, although a positive tendency was visible towards increased bone formation over time. Individual analyses by Gruen zones revealed that some zones, namely the proximal ones (e.g. G6), exhibit a broader scattering than others over time. This could be explained by the design of the evaluated tapered revision stem that achieves distal fixation and allows for proximal micromotion. Correlation analysis with subjective ratings (inter-rater reliability ICC = 0.71) showed a positive correlation with objective results, suggesting a feasibility of the method for clinical use.

CONCLUSIONS

In conclusion the presented method is an easy and accessible tool to quantify relative bone density changes during THA follow-up. It shows a positive correlation to established subjective assessment of bone remodelling and may therefore serve as a quantitative supplement in clinical documentation.