Conference Agenda

Introduction

In orthopaedics, traditional treatments generally involve the use of techniques and medical devices developed on the basis of population samples. However, they are unfortunately unlikely to account for the actual multiple characteristics specific to each individual patient, including all morphological and functional parameters of the anatomical complex under treatment. As surgical instrumentation and implantable devices are generally not sized to the patient's specific anatomy, which can lead to inaccurate implant positioning, impaired postoperative functional performance, and potentially to the surgical failure. The extraordinary advances in medical-imaging techniques and the advent of 3D printing, using basic biocompatible materials, i.e. polymeric/metal filament or powders, makes it possible to manufacture unique, highly complex medical devices. Hence, it is now possible and accessible to design and manufacture patient-specific surgical instrumentation and implant components that are well adapted to the patient's actual morphology. The aim of this study is to provide an in-depth, orthopaedic-focused view of the real potential and applications of 3D printing in the field of personalized surgery.

Advancements and Applications

Innovative solutions for the definition of fully customized medical devices for clinical applications have recently been developed at the Rizzoli Orthopaedic Institute in Bologna, Italy. These can be applied in conservative treatments or surgical procedures, these including minimal and massive skeletal reconstructions. A number of applications have been conducted in preclinical investigations and early clinical cases. Typically, they are based on reliable and accurate anatomical modelling using state-of-the-art medical imaging systems for morphological reconstruction. More specifically, these now include innovative weight-bearing CT based on cone-beam technology and EOS imaging, both with reduced radiation dose compared to standard devices and therefore safer for the patients and in paediatric applications, as well as dual-energy CT and 3.0T MRI supported by machine-learning protocols.

Preclinical applications included in-vitro testing of novel joint prostheses prototypes, investigation of lattice structures for bone growth into implant using cell cultures, sensitivity analysis on anatomical reconstruction from medical-image segmentation, and identification of the most appropriate clinical-oriented 3D printing parameters. The most recent clinical experiences include customized total knee and ankle replacement, personalised femur and tibia osteotomy for correction of varus-valgus deformity of the knee and ankle, massive pelvis reconstruction in oncologic patients, of spine deformity correction, vertebrectomy with subsequent reconstruction, patient-specific tibial intercalary segmental reconstruction, coronoid fracture repair, and many others. In all these applications, starting from CT/MRI-based morphological reconstructions of the patient's anatomy under treatment, the customised procedures involve surgical planning, design and EBM/SLM-based 3D printing, using biocompatible metal alloy powders, of the implantable devices and of the necessary surgical instrumentation. Other experiences have led to 3D printing of bio-models, which have proven to be essential in medical training, for preoperative planning and device preparation. The basic experiences principles can ideally be transferred to other medical branches.

Concluding Remarks

3D printing now seems clearly essential for the customization of orthopaedic treatments. We are confident that this technique will optimize patient-to-doctor communication, final patient performance, and medical training, resulting in cost-effective solutions for health care systems and industries.

Bone fractures are a common and a major cause of disability and death worldwide, and up to 10% of all fractures fail to heal normally (called “fracture non-unions”). Ultrasound imaging is safe, cost-effective, compatible with metal implants, and widely used to diagnose soft tissue injuries. Importantly, it holds significant potential to revolutionize fracture care by enabling early and routine monitoring and assessment of fracture healing, which for patients at high-risk of non-union, could mean earlier detection and clinical intervention of poor healing fractures that would prevent advanced clinical state and prolonged patient suffering, and reduce extensive waiting times for treatment, which would reduce costs for the healthcare system and patients. However, as ultrasound imaging is not routinely used in orthopedic clinics, barriers include difficulty acquiring and interpreting ultrasound images, lack of standardized scanning guidance and unclear terminology for bone and clinically important features of fracture healing. Previous efforts to use ultrasound imaging for musculoskeletal applications have demonstrated large intra‐examiner and inter‐examiner reliability variance for semi-quantitative and continuous measures, and image processing was slow as segmentation is often manual (or semi-automated). Therefore, there exists strong need to develop automated approaches to reduce the observed variance and to automate the segmentation procedure. We have established terminology and reporting guidelines for bone healing on ultrasound imaging; including expert agreement on key bone healing features from orthopaedic surgeons, radiographers and medical physics experts, and we have developed a Python-based, multi-label classification machine learning (ML) algorithm using these consensus labels to aid clinical interpretation of ultrasound scans. We have assess 2D US image label classification performance of four pre-trained Convolutional Neural Networks (CNNs) that have previously demonstrated good bone segmentation ability within US imaging in recent studies (VGG-16, VGG-19, Resnet50 and deeplabv3) using a dataset consisting of 734 2D ultrasound images of two superficial and frequently fractured long bones, the tibia (3 x longitudinal scan sweeps, 366 2D images) and the clavicle (3 x longitudinal scan sweeps, 368 2D images) acquired using an ultrasound scanner (L15 HD3, Clarius, Vancouver, BC, Canada). While all four tested algorithms performed with relatively high precision, recall/sensitivity and F1 scores, the deeplabv3 algorithm demonstrated the best segmentation performance across all metric parameters for bone segmentation on 2D US imaging, followed closely by Resnet50, then VGG-19 and VGG-16. Augmenting the dataset resulted in notable improvements to mean class accuracy particularly for VGG-16 and deeplabv3 and marginally decreased performance for VGG-19 and Resnet50. Overall, dataset augmentation resulted in marginal overall improvements. This is the first study to compare the performance of four different and state-of-the-art clinically used CNNs to identify the algorithm demonstrating the best segmentation performance for bone on US imaging. The results presented are significant for inter-disciplinary biomedical engineers to design efficient, accurate and robust automated ML-based segmentation models for clinical orthopedic applications.

Muscle strength assessment is a fundamental outcome measure for evaluating shoulder function in patients with musculoskeletal disorders. Hand-held dynamometers (HHDs) are widely used in clinical practice due to their ease of use, cost-effectiveness, and portability. However, factors such as examiner strength, subject positioning, and HHD placement often compromise their reliability. To address these limitations, recent advancements in orthopedic bioengineering have explored novel measurement devices, including load cells, which provide higher accuracy and improved data consistency.

This study proposed an innovative system that integrates an HHD with a load cell mounted on a rigid, 3D-printed, modular support, designed to reduce operator dependency and enhance measurement precision. By standardizing force measurements and eliminating variability due to manual positioning and tester influence, this configuration aims to improve the reproducibility of muscle strength assessments. Furthermore, the system enables experimental trials in which the load cell and HHD function in a serial configuration, allowing for a direct comparison of their precision and reliability in force measurement.

The system was validated through static weight tests ranging from 9.81 to 98.10 N, based on previous experiments conducted with healthy volunteers and literature studies involving patients with musculoskeletal shoulder disorders. Validation tests using static weights demonstrated good consistency between the two devices, though discrepancies became more pronounced as the weight increased. A feasibility assessment was also conducted with healthy volunteers, who underwent standardized strength evaluations under controlled conditions to assess the system's reliability and validity. Results confirmed the system's capability to measure shoulder muscle strength and facilitate comparisons across different measurement techniques. Specifically, the Bland-Altman analysis indicated a small systematic bias (mean difference: 1.6 N) and narrow Limits of Agreement (LOA: -3.8 N to 7.0 N). The Mean Absolute Error (MAE: 2.9 N) and Mean Absolute Percentage Error (MAPE: 9.4%) further demonstrated acceptable error levels in both dominant and non-dominant arms.

The integration of advanced bioengineering solutions into muscle strength assessment represents a significant step toward enhancing the accuracy, reliability, and standardization of shoulder function evaluation. By minimizing operator dependency and optimizing measurement protocols, this approach offers a robust framework for clinical assessments, rehabilitation monitoring, and postoperative outcome evaluations. Additionally, it offers a more objective and data-driven approach for evaluating musculoskeletal impairments, facilitating early diagnosis and personalized treatment planning. Future large-scale studies are needed to validate the system across diverse populations and clinical environments, ensuring its effectiveness in real-world applications. The proposed system demonstrates strong potential for clinical adoption, particularly in the continuous monitoring and diagnosis of musculoskeletal shoulder disorders, where precise and reproducible strength measurement is critical for guiding therapeutic interventions and optimizing patient outcomes.

Acknowledgement: Funded by the European Union - Next Generation EU - NRRP M6C2 - Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MAD-2022-12376080 - CUP: F83C22002450001).

Knee osteoarthritis (KOA) is a common degenerative musculoskeletal condition among the elderly, often resulting in chronic pain and reduced mobility. Numerous studies have shown that KOA significantly alters knee biomechanics, leading to changes in movement patterns. Therefore, analyzing lower limb biomechanics is critical for identifying parameters that describe functional limitations and quantifying the effectiveness of therapeutic interventions. To date, accurately characterizing the complex kinematics of the knee joint in patients with KOA remains a challenge. Optoelectronic motion capture systems, currently considered the gold standard, can provide quantitative assessments of 3D kinematics. In particular, gait analysis plays a crucial role in providing a detailed understanding of changes in movement and joint mechanics in subjects with KOA.

Over the past decades, a wide range of biomechanical models have been developed to estimate joint kinematics. Several comparative studies have been conducted on healthy populations to assess the performance of different gait models and marker configurations. These investigations have yielded valuable insights into the reliability and repeatability of kinematic outputs. However, limited attention has been given to evaluating such models in patients with KOA, who often present with knee deformities and soft tissue swelling, which can affect marker positioning and joint center estimation. Consequently, there remains a need to investigate the agreement between different models, specifically in this population, where such anatomical variations can increase uncertainty in kinematic results and impact clinical interpretation.

This study aim to investigate the inter-model agreement between two widely implemented gait models: the anatomically-based Istituto Ortopedico Rizzoli (IOR) model and the cluster-based Calibrated Anatomical System Technique (CAST) model, using both six degrees of freedom (6DoF) and inverse kinematics (IK) computational approaches.

A total of 16 patients was enrolled in this study. Each patient completed six walking trials on a 3-meter walkway at their own pace, with both the IOR and CAST marker sets applied simultaneously. Agreement between models was then quantitatively assessed using intraclass correlation coefficients (ICCs) and mean absolute differences (MD).

Results indicated high consistency across all models in measuring sagittal plane knee dynamics, with ICC values consistently exceeding 0.75 and mean absolute differences below 2.4°, demonstrating reliability during all phases of gait (heel strike, toe-off, loading response, terminal stance, and maximum swing). However, considerable variability emerged in frontal and transverse plane movements. Notably, frontal plane analyses revealed mean absolute differences as large as 12.7° during the maximum swing phase between 6DoF configurations of CAST and IOR models. Comparatively smaller discrepancies (≤5.8°) were noted between IK configurations.

The variations observed between IOR and CAST models, as well as between the 6DoF and IK computational methods in the frontal and transverse planes, highlighted the importance of refinement and validation of gait analysis methodologies. Careful model selection is crucial in clinical and research settings, particularly for populations with complex and variable joint dynamics like individuals with KOA.

Acknowledgment: Funded by the European Union - Next Generation EU-NRRP M6C2-Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MCNT2-2023-12378237-CUP: F87G24000130006)