Dataset Information

Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning.

ABSTRACT: Dual-energy X-ray absorptiometry (DXA) is the gold standard imaging method for diagnosing osteoporosis in clinical practice. The DXA images are commonly used to estimate a numerical value for bone mineral density (BMD), which decreases in osteoporosis. Low BMD is a known risk factor for osteoporotic fractures. In this study, we used deep learning to identify lumbar scoliosis and structural abnormalities that potentially affect BMD but are often neglected in lumbar spine DXA analysis. In addition, we tested the approach's ability to predict fractures using only DXA images. A dataset of 2949 images gathered by Kuopio Osteoporosis Risk Factor and Prevention Study was used to train a convolutional neural network (CNN) for classification. The model was able to classify scoliosis with an AUC of 0.96 and structural abnormalities causing unreliable BMD measurement with an AUC of 0.91. It predicted fractures occurring within 5 years from the lumbar spine DXA scan with an AUC of 0.63, meeting the predictive performance of combined BMD measurements from the lumbar spine and hip. In an independent test set of 574 clinical patients, AUC for lumbar scoliosis was 0.93 and AUC for unreliable BMD measurements was 0.94. In each classification task, neural network visualizations indicated the model's predictive strategy. We conclude that deep learning could complement the well established DXA method for osteoporosis diagnostics by analyzing incidental findings and image reliability, and improve its predictive ability in the future.

SUBMITTER: Nissinen T

PROVIDER: S-EPMC8102403 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ObjectivesFracture liaison services are designed to identify patients needing osteoporosis treatment after a fracture. Some fracture liaison service designs involve a prescreening step, for example, fracture risk assessment tool (FRAX®). Another possible prescreening tools are bone mass density assessment in the acute setting. The aim of this study was to assess the effectiveness of prescreening tools.MethodsIn the present prospective cohort study, women aged >55 years with a radius fracture were included. Patients were recruited at the emergency department after experiencing their fracture. All patients performed fracture risk assessment by fracture risk assessment tool, and bone mass density assessment by digital X-ray radiogrammetry and dual-energy X-ray absorptiometry (prescreening steps) as well as full routine evaluation at the osteoporosis unit (endpoint). The main outcome measures were sensitivity, specificity, predictive values, and area under the curve.ResultsForty-one women were recruited (mean age: 70 ± 8 years). Of these, 54% fulfilled the treatment indication criteria of osteoporosis after a full examination. Fracture risk assessment tool without bone mass density (cutoff ⩾ 15%) for prescreening patients had a high sensitivity (90%) but a low area under the curve (0.50) and specificity (16%). The highest area under the curve (0.73) was found prescreening with bone mass density assessment (dual-energy X-ray absorptiometry or digital X-ray radiogrammetry) having a sensitivity of 59%-86% and specificity of 61%-90%.ConclusionThis study, though small, raises questions regarding the effectiveness of using a prescreening step in fracture liaison services for high-risk individuals. In this cohort, FRAX® without bone mass density had a low precision, with a risk of both underestimating and overestimating patients requiring treatment. Bone mass density assessment in the acute setting could improve the precision of prescreening. Further investigations on the effectiveness and health economics of prescreening steps in fracture liaison services are needed.

Dataset Information

Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets