Related Experiment Video
Updated: Jan 15, 2026
09:02
Cortical Bone Assessment Using Ultrasonic Guided Waves: A Reproducibility Study in a Healthy Population
Published on: January 31, 2025
1.5K
Evaluating bone density with ultrasonic backscatter: Leveraging time-frequency analyses and convolutional neural
Hugh E Ferguson1, Carl D Herickhoff2, Ann M Viano1
1Department of Physics, Rhodes College, Memphis, TN 38112, USA.
Ultrasonics
|October 11, 2025
Summary
A new convolutional neural network (CNN) method uses ultrasound bone imaging to detect osteoporosis. This AI approach accurately predicts bone density, offering a sensitive tool for identifying osteoporotic changes.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Artificial Intelligence
Background:
- Osteoporosis is a condition of reduced bone density, increasing fracture risk.
- Current diagnostic methods for osteoporosis can be invasive or lack sensitivity.
- Advanced signal processing and machine learning offer potential for improved bone density assessment.
Purpose of the Study:
- To develop and validate a convolutional neural network (CNN) model for detecting bone density changes.
- To assess the efficacy of using spectrograms and scalograms from ultrasonic backscatter signals for bone density prediction.
- To compare the performance of the CNN method against conventional ultrasonic techniques.
Main Methods:
- Generated spectrograms and scalograms from ultrasonic backscatter signals of 55 human proximal femur cancellous bone specimens.
- Optimized CNN hyperparameters (learning rate, batch size, epochs) using grid search.
- Trained and tested the CNN model for bone density prediction and compared results with actual specimen densities.
Main Results:
- The CNN model accurately predicted bone density from ultrasound data, achieving high correlations (R²=0.98 for spectrograms, R²=0.94 for scalograms).
- Linear regression showed strong agreement between predicted and actual bone densities, with slopes near one.
- Spectrogram-based input slightly outperformed scalogram-based input for the CNN model.
- The CNN model demonstrated robust performance, with minimal sensitivity to hyperparameter tuning or training set size.
Conclusions:
- CNN analysis of ultrasound-derived spectrograms/scalograms provides a sensitive method for detecting osteoporotic bone changes.
- This AI-driven approach shows potential to outperform conventional ultrasonic backscatter methods for bone density assessment.
- The developed CNN method offers a promising, non-invasive technique for osteoporosis diagnosis and monitoring.
Related Concept Videos
01:17Ultrasonography
7.3K
Ultrasonography is an imaging technique that uses high-frequency sound waves to visualize the body's internal structures. It is a non-invasive and safe procedure that does not involve the use of ionizing radiation, making it widely used in various medical fields. Ultrasonography is used to study heart function, blood flow in the neck or extremities, certain conditions such as gallbladder disease, and fetal growth and development.
During an ultrasonography procedure, a handheld device called...
During an ultrasonography procedure, a handheld device called...
7.3K
01:18Classification of Bones
9.6K
The bones of the human skeletal system are of varied shapes, sizes, and functions. They can be classified based on their shape and function into four major classes: long bones, short bones, flat bones, and irregular bones. Some classifications include a fifth type, the sesamoid bones, as a separate class, whereas others categorize them under short bones.
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The...
9.6K