Background Clinical assessment of low back pain provides limited information on posture and movement behaviour during daily life. Wearable sensors can quantify free-living lumbar kinematics. This case series explored whether real-world lumbar posture differed between a day recorded before and two weeks after a personalised physiotherapy session, and what behaviour preceded naturally occurring pain on each day. Methods Five individuals with chronic, movement-evoked back pain were recruited. Lumbar posture was recorded with wearable sensors during a full day before and two weeks after a physiotherapy assessment and individualised postural advice, delivered by a physiotherapist blinded to sensor data, with pain events recorded in a diary. Within each participant, posture distributions were compared between days using the Wasserstein distance with block permutation tests, and time towards end-range flexion in the 5–15 min before pain was compared with random pain-free periods from the same day. Findings Whole-day posture differed between days in three participants, alongside marked differences in activity. Time towards end-range flexion changed in the direction of the postural advice in four of five participants. Pain was preceded by more time towards end-range flexion in three participants, matching the clinical assessment in two. Pain events fell in four participants, and all five reported reduced disability. Interpretation In this small case series, wearable sensors captured individual differences in real-world postural behaviour and in the behaviour preceding pain, which did not always match the clinical assessment. Without repeated baseline measurements, differences cannot be attributed to the intervention, and the findings are hypothesis-generating.
Background Temporomandibular joint (TMJ) biomechanics can be characterized by mandibular motion, masticatory muscle activity, and bite force generation. These measurements can provide functional inputs for subject-specific computational analyses of internal joint mechanics. However, existing tools typically measure these signals using separate hardware and software platforms, complicating their combined acquisition at chairside. Methods We developed and technically evaluated a multimodal platform for chairside acquisition of TMJ biomechanical data. The platform integrates motion, bite force, muscle activity, acoustic, and event-timing measurements with software for real-time preview, protocol guidance, and synchronized export. Technical evaluation included rigid-bar tracking tests, spectacle-frame stability testing during jaw tasks, and static force-sensor linearity testing. We analyzed pre- and postoperative data from an orthognathic surgery patient using offline kinematic, force-control, and computational modeling workflows. Findings Rigid-bar tracking tests yielded mean length deviations of 0.03 mm under static conditions and mean RMSE of 0.16 mm during movement. Spectacle-to-maxillary RMS translational and rotational differences were 0.28 ± 0.04 mm and 0.06° (SD < 0.01°) during maximum mouth opening-closing, and 0.30 ± 0.09 mm and 0.13 ± 0.03° during dynamic biting. Linear regression of force-sensor measurements against applied loads of 0–50 N yielded R2 = 0.998. In the illustrative case, platform setup took approximately 15 min, followed by approximately 15 min of chairside data acquisition. Offline analysis yielded pre- and postoperative mandibular kinematic measurements, bite-force tracking deviations, and model-estimated TMJ disc stresses. Interpretation The platform enabled chairside acquisition of multimodal biomechanical data for subsequent offline analysis. This platform could support subject-specific biomechanical analysis and future clinical studies of TMJ function.
Background Pediatric posterior cruciate ligament (PCL) biomechanics remain poorly characterized, and validated computational models are limited. The purpose of this study was to evaluate whether pediatric knee finite element models could reproduce intact cadaveric posterior tibial translation under PCL-dominant loading. Methods Twenty-two pediatric knee models were generated from computed tomography scans using a previously established juvenile modeling workflow. Model predictions were benchmarked against published intact cadaveric data for two loading conditions: a 134 N posterior tibial load at 30° flexion and a coupled 100 N posterior tibial load with 5 Nm external rotation at 0° of flexion. Posterior tibial translation was compared using Welch's two-sample t-tests and confidence intervals. Findings Under isolated posterior loading at 30°, the models predicted 9.4 ± 1.4 mm of posterior tibial translation compared with 10.8 ± 3.7 mm in the cadaveric benchmark (p = 0.15). No significant difference was found under the coupled posterior loading with external rotation between the models predicted posterior translations of 6.9 ± 0.6 mm and the experimental translation at 6.5 ± 3.1 (p = 0.56). Interpretation These findings support the use of specimen-specific pediatric knee models as an externally benchmarked platform for studying PCL biomechanics. Accurate reproduction of cadaveric posterior laxity under clinically relevant loading provides a foundation for future analyses of pediatric injury mechanisms and surgical and rehabilitation strategies.
