Bone Mineral Density Prediction from CT Image: A Novel Approach using ANN

Area of Science:

• Biomedical Imaging
• Artificial Intelligence in Medicine
• Osteoporosis Research

Background:

• Osteoporosis remains underdiagnosed and undertreated, despite being treatable.
• Bone mineral density (BMD) monitoring is crucial for predicting and preventing osteoporosis-related complications.
• Quantitative computed tomography (QCT) measures BMD but doesn't fully utilize bone architecture, which is vital for aging populations.

Purpose of the Study:

• To develop an innovative, cost-effective method for predicting BMD by incorporating bone architecture.
• To leverage existing clinical CT scans for BMD assessment, reducing additional costs, time, and radiation exposure.

Main Methods:

• An artificial neural network (ANN) model was employed to predict BMD from clinical CT scan images.
• Image processing techniques extracted relevant bone properties from DICOM images.
• A backpropagation neural network with five input neurons and a hidden layer of 40 neurons was utilized.

Main Results:

• The ANN model achieved a high correlation coefficient of 0.883 between predicted BMD and QCT-derived density values.
• The model successfully predicted bone density using image properties from clinical CT scans of rabbit femur bones.
• The approach demonstrated the feasibility of predicting BMD without additional scans or radiation.

Conclusions:

• The proposed ANN-based approach offers a non-invasive and cost-effective method for BMD prediction.
• This technique can aid clinicians in the early identification of osteoporosis.
• It provides a pathway for developing strategies to improve BMD and manage osteoporosis effectively.