Sub-millimeter quantification of alveolar bone loss using automated 40 MHz high-frequency ultrasound: A proof-of-concept ex vivo validation study

This proof-of-concept study evaluated the performance of high-frequency ultrasound (HFUS, 40 MHz) for monitoring changes in alveolar bone loss and determining its detection capability and measurement precision. The study aimed to quantify system error, minimum detectable change, and agreement with a photographic ground truth measurement. An ex vivo model was used in which controlled bone loss increments were created and measured using both manual and automated HFUS methods. Measurements were compared to calibrated photographic measurements as the reference standard. System error was calculated using repeated baseline measurements, and detection capability was assessed using sensitivity analysis across incremental bone loss values. Agreement between methods was evaluated using Bland–Altman analysis and intraclass correlation coefficients (ICC). The system error ranged from 77 µm for automated image-registration measurements to 113 µm for manual measurements. The minimum detectable bone loss increment was experimentally determined to be approximately 138 µm. Detection sensitivity ranged from 79% to 100%, depending on the measurement method and increment size. Automated ultrasound measurements showed strong agreement for longitudinal change tracking, with a best-case bias of −18 µm, corresponds to less than 10% error relative to a typical 200 µm bone change increment and excellent reliability (ICC=0.89–0.99). No significant systematic bias was observed compared with photographic measurements (P>0.05), indicating good agreement between methods. High-frequency ultrasound demonstrated the ability to detect small longitudinal changes in alveolar bone loss and to track bone loss progression over time with high measurement precision. The main contribution of this study is demonstrating that HFUS can be used as a radiation-free method for monitoring longitudinal changes in alveolar bone loss rather than relying solely on absolute measurement accuracy.