Machine learning for automated abdominal aortic calcification scoring of DXA vertebral fracture assessment images: A pilot study

Quantifying the calcification of abdominal aorta and major side branches with deep learning

AIM

To explore the possibility of a neural network-based method for quantifying calcifications of the abdominal aorta and its branches.

MATERIALS AND METHODS

In total, 58 computed tomography (CT) angiography volumes were selected from a dataset of 609 to represent different stages of sclerosis. The ground truth segmentations of the abdominal aorta, coeliac trunk, superior mesenteric artery, renal arteries, common iliac arteries, and their calcifications were delineated manually. Two V-Net ensemble models were trained, one for segmenting arteries of interest and another for calcifications. The branches of interest were shortened algorithmically. The volumes of calcification were then evaluated from the arteries of interest.

RESULTS

The results indicate that automatic detection is possible with a high correlation to the ground truth. The scores for the ensemble calcification model were dice score of 0.69 and volumetric similarity (VS) of 0.80 and for the arteries of interest segmentations: aorta: dice 0.96, VS 0.98; aortic branches: dice 0.74, VS 0.87; and common iliac arteries: dice 0.72, VS 0.91.

CONCLUSIONS

The presented neural network model is the first to be capable of automatically segmenting, in addition to calcification, both the aorta and its branches from contrast-enhanced CT angiography. This technology shows promise in addressing limitations inherent in earlier methods that relied solely on plain CT.

Radiofrequency Echographic Multispectrometry (REMS) can Overcome the Effects of Structural Internal Artifacts and Evaluate Bone Fragility Accurately

PURPOSE

This study measured bone mineral density (BMD) in a Japanese population using the novel non-ionizing system using radiofrequency echographic multispectrometry (REMS) and compared the results with those obtained using traditional dual-energy X-ray absorptiometry (DXA). We aimed to identify any discrepancies between measurements obtained using these instruments and identify the influencing factors.

METHODS

This cross-sectional study examined patients with osteoporosis treated at a single center from April to August 2023. We examined BMD assessment by DXA and REMS in lumbar spine and proximal femur. Patients were categorized into two groups: those with discrepancies between lumbar spine BMD measured by DXA and REMS, and those without. Semiquantitative evaluation of vertebral fractures and abdominal aortic calcification scoring were also performed and compared between the two groups, along with various patient characteristics.

RESULTS

A total of 70 patients (88.6% female; mean age 78.39 ± 9.50 years) undergoing osteoporosis treatment were included in the study. A significant difference was noted between DXA and REMS measurement of BMD and T-scores, with REMS recording consistently lower values. The discrepancy group exhibited a higher incidence of multiple vertebral fractures and increased vascular calcification than the non-discrepancy group. Multivariate analysis indicated that diabetes mellitus, severe vertebral fractures, and increased abdominal aortic calcification scores were significantly associated with discrepancies in lumbar spine T-scores.

CONCLUSION

This study suggests that REMS may offer a more accurate measurement of BMD, overcoming the overestimation of BMD by DXA owing to factors such as vertebral deformities, abdominal aortic calcification, and diabetes mellitus.

Extent of Abdominal Aortic Calcification Is Associated With Incident Rapid Weight Loss Over 5 Years: The Perth Longitudinal Study of Ageing Women

BACKGROUND

Abdominal aortic calcification (AAC), a marker of vascular disease, is associated with disease in other vascular beds including gastrointestinal arteries. We investigated whether AAC is related to rapid weight loss over 5 years and whether rapid weight loss is associated with 9.5-year all-cause mortality in community-dwelling older women.

METHODS

Lateral spine images from dual-energy x-ray absorptiometry (1998/1999) were used to assess AAC (24-point AAC scoring method) in 929 older women. Over 5 years, body weight was assessed at 12-month intervals. Rapid weight loss was defined as >5% decrease in body weight within any 12-month interval. Multivariable-adjusted logistic regression was used to assess AAC and rapid weight loss and Cox regression to assess the relationship between rapid weight loss and 9.5-year all-cause mortality.

RESULTS

Mean±SD age of women was 75.0±2.6 years. During the initial 5 years, 366 (39%) women presented with rapid weight loss. Compared with women with low AAC (24-point AAC score 0-1), those with moderate (24-point AAC score 2-5: odds ratio, 1.36 [95% CI, 1.00-1.85]) and extensive (24-point AAC score 6+: odds ratio, 1.59 [95% CI, 1.10-2.31]) AAC had higher odds for presenting with rapid weight loss. Results remained similar after further adjustment for dietary factors (alcohol, protein, fat, and carbohydrates), diet quality, blood pressure, and cholesterol measures. The estimates were similar in subgroups of women who met protein intake (n=599) and physical activity (n=735) recommendations (extensive AAC: odds ratios, 1.81 [95% CI, 1.12-2.92] and 1.58 [95% CI, 1.02-2.44], respectively). Rapid weight loss was associated with all-cause mortality over the next 9.5 years (hazard ratio, 1.49 [95% CI, 1.17-1.89]; P=0.001).

CONCLUSIONS

AAC extent was associated with greater risk for rapid weight loss over 5 years in older women, a risk for all-cause mortality. Since the association was unchanged after taking nutritional intakes into account, these data support the possibility that vascular disease may play a role in the maintenance of body weight.

Machine-Learning Assessed Abdominal Aortic Calcification is Associated with Long-Term Fall and Fracture Risk in Community-Dwelling Older Australian Women

Abdominal aortic calcification (AAC), a recognized measure of advanced vascular disease, is associated with higher cardiovascular risk and poorer long-term prognosis. AAC can be assessed on dual-energy X-ray absorptiometry (DXA)-derived lateral spine images used for vertebral fracture assessment at the time of bone density screening using a validated 24-point scoring method (AAC-24). Previous studies have identified robust associations between AAC-24 score, incident falls, and fractures. However, a major limitation of manual AAC assessment is that it requires a trained expert. Hence, we have developed an automated machine-learning algorithm for assessing AAC-24 scores (ML-AAC24). In this prospective study, we evaluated the association between ML-AAC24 and long-term incident falls and fractures in 1023 community-dwelling older women (mean age, 75 ± 3 years) from the Perth Longitudinal Study of Ageing Women. Over 10 years of follow-up, 253 (24.7%) women experienced a clinical fracture identified via self-report every 4-6 months and verified by X-ray, and 169 (16.5%) women had a fracture hospitalization identified from linked hospital discharge data. Over 14.5 years, 393 (38.4%) women experienced an injurious fall requiring hospitalization identified from linked hospital discharge data. After adjusting for baseline fracture risk, women with moderate to extensive AAC (ML-AAC24 ≥ 2) had a greater risk of clinical fractures (hazard ratio [HR] 1.42; 95% confidence interval [CI], 1.10-1.85) and fall-related hospitalization (HR 1.35; 95% CI, 1.09-1.66), compared to those with low AAC (ML-AAC24 ≤ 1). Similar to manually assessed AAC-24, ML-AAC24 was not associated with fracture hospitalizations. The relative hazard estimates obtained using machine learning were similar to those using manually assessed AAC-24 scores. In conclusion, this novel automated method for assessing AAC, that can be easily and seamlessly captured at the time of bone density testing, has robust associations with long-term incident clinical fractures and injurious falls. However, the performance of the ML-AAC24 algorithm needs to be verified in independent cohorts. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Automatic Segmentation and Quantification of Abdominal Aortic Calcification in Lateral Lumbar Radiographs Based on Deep-Learning-Based Algorithms

To investigate the performance of deep-learning-based algorithms for the automatic segmentation and quantification of abdominal aortic calcification (AAC) in lateral lumbar radiographs, we retrospectively collected 1359 consecutive lateral lumbar radiographs. The data were randomly divided into model development and hold-out test datasets. The model development dataset was used to develop U-shaped fully convolutional network (U-Net) models to segment the landmarks of vertebrae T12-L5, the aorta, and anterior and posterior aortic calcifications. The AAC lengths were calculated, resulting in an automatic Kauppila score output. The vertebral levels, AAC scores, and AAC severity were obtained from clinical reports and analyzed by an experienced expert (reference standard) and the model. Compared with the reference standard, the U-Net model demonstrated a good performance in predicting the total AAC score in the hold-out test dataset, with a correlation coefficient of 0.97 (p <0.001). The overall accuracy for the AAC severity was 0.77 for the model and 0.74 for the clinical report. Additionally, the Kendall coefficient of concordance of the total AAC score prediction was 0.89 between the model-predicted score and the reference standard, and 0.88 between the structured clinical report and the reference standard. In conclusion, the U-Net-based deep learning approach demonstrated a relatively high model performance in automatically segmenting and quantifying ACC.

Machine learning for abdominal aortic calcification assessment from bone density machine-derived lateral spine images

BACKGROUND

Lateral spine images for vertebral fracture assessment can be easily obtained on modern bone density machines. Abdominal aortic calcification (AAC) can be scored on these images by trained imaging specialists to assess cardiovascular disease risk. However, this process is laborious and requires careful training.

METHODS

Training and testing of model performance of the convolutional neural network (CNN) algorithm for automated AAC-24 scoring utilised 5012 lateral spine images (2 manufacturers, 4 models of bone density machines), with trained imaging specialist AAC scores. Validation occurred in a registry-based cohort study of 8565 older men and women with images captured as part of routine clinical practice for fracture risk assessment. Cox proportional hazards models were used to estimate the association between machine-learning AAC (ML-AAC-24) scores with future incident Major Adverse Cardiovascular Events (MACE) that including death, hospitalised acute myocardial infarction or ischemic cerebrovascular disease ascertained from linked healthcare data.

FINDINGS

The average intraclass correlation coefficient between imaging specialist and ML-AAC-24 scores for 5012 images was 0.84 (95% CI 0.83, 0.84) with classification accuracy of 80% for established AAC groups. During a mean follow-up 4 years in the registry-based cohort, MACE outcomes were reported in 1177 people (13.7%). With increasing ML-AAC-24 scores there was an increasing proportion of people with MACE (low 7.9%, moderate 14.5%, high 21.2%), as well as individual MACE components (all p-trend <0.001). After multivariable adjustment, moderate and high ML-AAC-24 groups remained significantly associated with MACE (HR 1.54, 95% CI 1.31-1.80 & HR 2.06, 95% CI 1.75-2.42, respectively), compared to those with low ML-AAC-24.

INTERPRETATION

The ML-AAC-24 scores had substantial levels of agreement with trained imaging specialists, and was associated with a substantial gradient of risk for cardiovascular events in a real-world setting. This approach could be readily implemented into these clinical settings to improve identification of people at high CVD risk.

FUNDING

The study was supported by a National Health and Medical Research Council of Australia Ideas grant and the Rady Innovation Fund, Rady Faculty of Health Sciences, University of Manitoba.

Association between a body shape index and abdominal aortic calcification in general population: A cross-sectional study

Background

The association between a body shape index (ABSI) and abdominal aortic calcification (AAC) is still unclear, so we tried to prove the association between ABSI and AAC in the general population in this cross-sectional study.

Materials and methods

After excluding participants with missing data on height, weight, waist circumference (WC), and AAC, we finally selected 3,140 participants aged 40-80 years from the 2013-2014 National Health and Nutrition Examination Survey. Using multivariate logistic regression and receiver operating characteristic (ROC) curves to test the association between ABSI and AAC.

Results

Participants (median age: 58.0 years; 48.3% men) were divided into two groups by the optimal cutoff point of ABSI: higher ABSI (> 0.84) and lower ABSI (≤ 0.84). Participants with higher ABSI showed significantly higher proportion of AAC than those with lower ABSI (39.8 vs. 23.7%, P < 0.001). Participants with higher ABSI had an increased risk of developing AAC in crude model (ABSI as a continuous variable: OR = 2.485, 95% CI: 2.099-2.942, P < 0.001; as a categorical variable: OR = 2.132, 95% CI: 1.826-2.489, P < 0.001), and ABSI was still independently associated with AAC in all adjusted models (all P < 0.05). Further subgroup analyses showed that higher ABSI was consistently associated with AAC in subgroups with sex (male or female), age (≤ 65 or > 65 years), smoking history (yes or no), hypertension (yes or no), diabetes (yes or no), sleep disorder (yes or no), body mass index (BMI) (< 23 or ≥ 23 kg/m²), systolic blood pressure (< 140 or ≥ 140 mmHg), diastolic blood pressure (< 90 or ≥ 90 mmHg), fasting plasma glucose (< 126 or ≥ 126 mg/dL), and low-density lipoprotein cholesterol (≤ 130 or > 130 mg/dL) (P for interaction > 0.05). While in other subgroups, the association was no longer synchronized. The ROC showed that the area under the curve of ABSI was significantly higher than height, weight, BMI, WC, and waist-to-height ratio (WHtR).

Conclusion

Higher ABSI was closely associated with higher risk of AAC, and discriminant ability of ABSI for AAC was significantly higher than height, weight, BMI, WC, and WHtR.

A Novel Quantitative Computer-Assisted Score Can Improve Repeatability in the Estimate of Vascular Calcifications at the Abdominal Aorta

In CKD and in the elderly, Vascular Calcifications (VC) are associated to cardiovascular events and bone fractures. VC scores at the abdominal aorta (AA) from lateral spine radiographs are widely applied (the 0-24 semiquantitative discrete visual score (SV) being the most used). We hypothesised that a novel continuum score based on quantitative computer-assisted tracking of calcifications (QC score) can improve the precision of the SV score. This study tested the repeatability and reproducibility of QC score and SV score. In forty-four patients with VC from an earlier study, five experts from four specialties evaluated the data twice using a dedicated software. Test-retest was performed on eight subjects. QC results were reported in a 0-24 scale to readily compare with SV. The QC score showed higher intra-operator repeatability: the 95% CI of Bland-Altman differences was almost halved in QC; intra-operator R2 improved from 0.67 for SV to 0.79 for QC. Inter-observer repeatability was higher for QC score in the first (Intraclass Correlation Coefficient 0.78 vs. 0.64), but not in the second evaluation (0.84 vs. 0.82), indicating a possible heavier learning artefact for SV. The Minimum Detectable Difference (MDD) was smaller for QC (2.98 vs. 4 for SV, in the 0-24 range). Both scores were insensitive to test-retest procedure. Notably, QC and SV scores were discordant: SV showed generally higher values, and an increasing trend of differences with VC severity. In summary, the new QC score improved the precision of lateral spine radiograph scores in estimating VC. We reported for the first time an estimate of MDD in VC assessment that was 25% lower for the new QC score with respect to the usual SV score. An ongoing study will determine whether this lower MDD may reduce follow-up times to check for VC progression.

Abdominal aortic calcification on lateral spine images captured during bone density testing and late-life dementia risk in older women: A prospective cohort study

BACKGROUND

Dementia after the age of 80 years (late-life) is increasingly common due to vascular and non-vascular risk factors. Identifying individuals at higher risk of late-life dementia remains a global priority.

METHODS

In prospective study of 958 ambulant community-dwelling older women (≥70 years), lateral spine images (LSI) captured in 1998 (baseline) from a bone density machine were used to assess abdominal aortic calcification (AAC). AAC was classified into established categories (low, moderate and extensive). Cardiovascular risk factors and apolipoprotein E (APOE) genotyping were evaluated. Incident 14.5-year late-life dementia was identified from linked hospital and mortality records.

FINDINGS

At baseline women were 75.0 ± 2.6 years, 44.7% had low AAC, 36.4% had moderate AAC and 18.9% had extensive AAC. Over 14.5- years, 150 (15.7%) women had a late-life dementia hospitalisation (n = 132) and/or death (n = 58). Compared to those with low AAC, women with moderate and extensive AAC were more likely to suffer late-life dementia hospitalisations (9.3%, 15.5%, 18.3%, respectively) and deaths (2.8%, 8.3%, 9.4%, respectively). After adjustment for cardiovascular risk factors and APOE, women with moderate and extensive AAC had twice the relative hazards of late-life dementia (moderate, aHR 2.03 95%CI 1.38-2.97; extensive, aHR 2.10 95%CI 1.33-3.32), compared to women with low AAC.

INTERPRETATION

In community-dwelling older women, those with more advanced AAC had higher risk of late-life dementia, independent of cardiovascular risk factors and APOE genotype. Given the widespread use of bone density testing, simultaneously capturing AAC information may be a novel, non-invasive, scalable approach to identify older women at risk of late-life dementia.

FUNDING

Kidney Health Australia, Healthway Health Promotion Foundation of Western Australia, Sir Charles Gairdner Hospital Research Advisory Committee Grant, National Health and Medical Research Council of Australia.

Cardiovascular Disease Diagnosis from DXA Scan and Retinal Images Using Deep Learning

Cardiovascular diseases (CVD) are the leading cause of death worldwide. People affected by CVDs may go undiagnosed until the occurrence of a serious heart failure event such as stroke, heart attack, and myocardial infraction. In Qatar, there is a lack of studies focusing on CVD diagnosis based on non-invasive methods such as retinal image or dual-energy X-ray absorptiometry (DXA). In this study, we aimed at diagnosing CVD using a novel approach integrating information from retinal images and DXA data. We considered an adult Qatari cohort of 500 participants from Qatar Biobank (QBB) with an equal number of participants from the CVD and the control groups. We designed a case-control study with a novel multi-modal (combining data from multiple modalities-DXA and retinal images)-to propose a deep learning (DL)-based technique to distinguish the CVD group from the control group. Uni-modal models based on retinal images and DXA data achieved 75.6% and 77.4% accuracy, respectively. The multi-modal model showed an improved accuracy of 78.3% in classifying CVD group and the control group. We used gradient class activation map (GradCAM) to highlight the areas of interest in the retinal images that influenced the decisions of the proposed DL model most. It was observed that the model focused mostly on the centre of the retinal images where signs of CVD such as hemorrhages were present. This indicates that our model can identify and make use of certain prognosis markers for hypertension and ischemic heart disease. From DXA data, we found higher values for bone mineral density, fat content, muscle mass and bone area across majority of the body parts in CVD group compared to the control group indicating better bone health in the Qatari CVD cohort. This seminal method based on DXA scans and retinal images demonstrate major potentials for the early detection of CVD in a fast and relatively non-invasive manner.