Prediction of osteoporosis using radiomics analysis derived from single source dual energy CT

Abstract

Background With the aging population of society, the incidence rate of osteoporosis is increasing year by year. Early diagnosis of osteoporosis plays a significant role in the progress of disease prevention. As newly developed technology, computed tomography (CT) radiomics could discover radiomic features difficult to recognize visually, providing convenient, comprehensive and accurate osteoporosis diagnosis. This study aimed to develop and validate a clinical-radiomics model based on the monochromatic imaging of single source dual-energy CT for osteoporosis prediction. Methods One hundred sixty-four participants who underwent both single source dual-energy CT and quantitative computed tomography (QCT) lumbar-spine examination were enrolled in a study cohort including training datasets (n = 114 [30 osteoporosis and 84 non-osteoporosis]) and validation datasets (n = 50 [12 osteoporosis and 38 non-osteoporosis]). One hundred seven radiomics features were extracted from 70-keV monochromatic CT images. With QCT as the reference standard, a radiomics signature was built by using least absolute shrinkage and selection operator (LASSO) regression on the basis of reproducible features. A clinical-radiomics model was constructed by incorporating the radiomics signature and a significant clinical predictor (age) using multivariate logistic regression analysis. Model performance was assessed by its calibration, discrimination and clinical usefulness. Results The radiomics signature comprised 14 selected features and showed good calibration and discrimination in both training and validation cohorts. The clinical-radiomics model, which incorporated the radiomics signature and a significant clinical predictor (age), also showed good discrimination, with an area under the receiver operating characteristic curve (AUC) of 0.938 (95% confidence interval, 0.903–0.952) in the training cohort and an AUC of 0.988 (95% confidence interval, 0.967–0.998) in the validation cohort, and good calibration. The clinical-radiomics model stratified participants into groups with osteoporosis and non-osteoporosis with an accuracy of 94.0% in the validation cohort. Decision curve analysis (DCA) demonstrated that the radiomics signature and the clinical-radiomics model were clinically useful. Conclusions The clinical-radiomics model incorporating the radiomics signature and a clinical parameter had a good ability to predict osteoporosis based on dual-energy CT monoenergetic imaging.

Artificial Intelligence Applications for Osteoporosis Classification Using Computed Tomography

“…Clinical data and demographic traits [160,161] have been employed to identify individuals with osteoporosis or osteoporotic fractures, playing a pivotal role in the development of various tools for osteoporosis assessment [162]. The inclusion of clinical characteristics (such as age, sex, and risk factors as per the National Osteoporosis Foundation Guidelines [145]) in BMD assessment using DEXA has previously shown promising results in predicting osteoporosis, as demonstrated by Wang et al [104]. Their radiomics clinical model achieved an AUC of 0.988, compared with an AUC of 0.902 with radiomics alone (although this difference was not statistically significant, p = 0.643).…”

“…These features capture intricate details in the images (beyond the human eye), enabling a more comprehensive understanding of the underlying tissue characteristics. Radiomics and texture analysis with deep learning techniques have also been used to analyze trabecular bone structure in CT images, providing insights into bone quality and microstructure that are not attainable through DEXA or visual measurements alone, thereby improving the accuracy of diagnosing osteoporosis [104,105]. While many studies have focused on the use of AI for classifying or detecting osteoporosis, they often suffer from limitations, such as single-center designs, limited patient samples, and a lack of validation in real clinical settings.…”

“…The reference BMD density values for these internal references are determined using phantomcalibrated scans from a cohort of patients [103]. These values are then extrapolated to create a standard calibration curve for converting the trabecular HU to BMD [104,105] through scan-specific equations.…”

Quantitative computed tomography assessment of bone mineral density in adolescent idiopathic scoliosis: correlations with Cobb angle, vertebral rotation, and Risser sign

“…Quantitative computed tomography (QCT) is an accurate measurement offering a sophisticated and precise method for BMD assessment, with cross-sectional imaging providing a detailed volumetric analysis ( 13 ). This technology is superior to dual-energy X-ray absorptiometry (DXA) in differentiating between trabecular and cortical bone, offering a three-dimensional perspective that is less likely to be confounded by extraneous factors such as spinal degeneration or aortic calcifications ( 14 ). A previous study conducted by Almomen et al demonstrated that in women diagnosed with AIS, those with more pronounced spinal curvature, as indicated by higher Cobb angle measurements, are more likely to have lower BMD values, as determined through DXA assessments ( 15 ).…”

What To Expect (and What Not) From Dual-energy CT Imaging Now and in the Future?

“…The clinical value of DECT-derived radio(geno)mics features has been reported for diagnosis of malignancy, depicting genetic, molecular, and histological features in different tumor types; evaluating tumor invasiveness; predicting tumor staging; characterization of malignant lymph nodes, predicting patient outcome and survival, and predicting and evaluating tumor response in different tumor types [81][82][83][84][85][86][87] [Figure 23]. There are also preliminary data that support the possible value of DECT-derived features as imaging biomarkers in other non-oncologic applications such as the diagnosis of pulmonary embolism, type 2 diabetes mellitus, osteoporosis, liver fibrosis, and carotid stenosis, the assessment of lung interstitial disease or Crohn´s disease activity, or acute pancreatitis, the characterization of adrenal masses, and the location and analysis of kidney stones [88][89][90][91][92][93][94][95][96] [Figure 24]. The robustness of features has been identified as pivotal for the clinical implementation.…”