.png){kind=icon}
Return
Multidirectional Optical Bone Densitometry Using a Simulation-Based Machine Learning Model: Experimental Validation with Bone Phantoms
.png){kind=icon}
OA
AI
DOI：10.1007/s10439-026-04258-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
While the multidirectional optical bone densitometry approach, originally proposed based on simulation, has been previously reported, its experimental validation has not yet been demonstrated. This study addresses this gap by validating the method using bone phantoms. These phantoms were fabricated with Intralipos, agarose, and bovine cortical bone fragments. By varying the mixing ratio of cortical bone fragments, phantoms with different bone densities were prepared. Near-infrared laser light with a wavelength of 850 nm was directed onto the phantoms from three directions, and the resulting backward, lateral, and forward light intensity distributions on the phantom surface were recorded. Directional light intensity values were extracted from these distributions and used as features for the simulation-based machine learning model to predict bone density. Linear regression between the predicted and reference bone densities yielded an r2 of 0.85. These findings demonstrate the feasibility of the proposed method as a proof-of-concept validation and suggest a possible direction for future development of optical bone density assessment methods.
Keywords:
Bone mineral density
Optical bone densitometry
Machine learning
Simulation
Osteoporosis screening
Journal
IF:
5.4
Papers:
6.3K
Citations:
1.4W
Organization
Cited Papers
ShareSaveShareSaveShareSaveShareSaveDevice-Simulation-Based Machine Learning Technique for the Characteristic of Line Tunnel Field-Effect Transistors.png)
IEEE ACCESSOAAI
IEEE ACCESSIF3.6
OAAIShareSaveBiomolecular simulation based machine learning models accurately predict sites of tolerability to the unnatural amino acid acridonylalanineSCIENTIFIC REPORTS
OAAI
SCIENTIFIC REPORTSIF3.9
OAAIShareSave
View more