Abstract
New designs for total hip replacement (THR) are continuously proposed to solve the limitations of the actual solutions. However, before they can reach the clinical practice, a long and expensive procedure is required, with experimental in vitro and in vivo tests. It is possible to simplify this procedure using an approach known as In Silico Trials. The aim of this project is to exploit this approach and develop a complete solution, based on in silico simulations, to quantitatively assess the risk associated with the most common failure modes of new designs for THR before any experimental study is conducted. The research activity is articulated in several steps. At first, a large collection of morpho-densitometric digital models of the hip joint is extracted from the already available dataset to represent well the anatomical variability typical of the patients who receive this kind of treatment. Then, on each model, a THR surgical procedure is simulated, reproducing the best practice, but also the variability normally seen in the real surgical procedures. In this way, a wide cohort is defined combining for each virtual patient multiple surgeries to account for the surgical variability for a total of about 300 case studies. For each case study, patient-specific and surgery-specific musculoskeletal dynamics and finite element models are generated and run with some different boundary conditions, including several motor tasks and accounting for tasks variability and patient specificity. Such simulation cohort represents a proper in silico trial of hip joint replacement to predict the risks of implant failure with respect to three failure modes, i.e. aseptic loosening for the stem, wear for the artificial joint, and impingement and dislocation. These represent 70-80% of the THR failures reported in different outcome registers, and thus provide an excellent starting point for the validation. The clinical validation of this in silico trials technology will be necessarily retrospective, using the results of the RIPO Joint Replacement Outcome Registry. The RIPO register is currently monitoring the outcome of 187,000 hip replacement procedures. From these, we will select a few implant designs that show incidence of the three considered failure modes much higher or much lower than the average. Among them, we will validate our IST4HR simulator using models for which the implant design is available from the manufacturer, in the public domain, or can be reverse-engineered from postoperative CT scans. We already have three designs available, two from Zimmer-Biomet and one in public domain. We will use our IST4HR simulator to predict the incidence of each failure mode considered for each implant design, and then compare it with the real incidence of that failure mode as reported by the RIPO registry. Once validated, the simulator can be extended to include more failure modes, and also other types of joint replacements. RESULTS ACHIEVED During the project duration, UNIBO carried out research activities focused on the development and implementation of an automated computational framework for the assessment of the primary stability of cementless femoral stems (WP2-3), addressing one of the most clinically relevant failure mechanisms in total hip arthroplasty and a core objective of the IST4HR project. To this aim, UNIBO first screened the large collection of patient data from the publicly available, open-access HFValid dataset, including femoral CT scans, 3D femur geometry reconstructions, and calibration phantom data for bone material mapping. Ten patients were selected, and a team of orthopaedists from UNICA, in collaboration with UNIBO, virtually simulated at least three surgical procedures for each virtual patient using an ad hoc–developed, multi-modal display interface for hip replacement 3D planning inspired by the Hip-Op software. A central line of activity concerned the integration of three-dimensional preoperative planning tools with subject-specific finite element modelling, enabling the generation of realistic virtual surgical scenarios starting from clinical imaging data [3]. This approach allowed an explicit representation of implant positioning, surgical resection and bone–implant mechanical interaction, providing a biomechanically consistent basis for the assessment of primary mechanical stability under physiologically relevant loading conditions. Particular emphasis was placed on automation and reproducibility. A fully automated computational pipeline was developed to manage geometry processing, Boolean operations, meshing, material property mapping and simulation execution with minimal manual intervention (Fig. 5). This automation significantly reduced model preparation time and enabled systematic analyses on virtual patient cohorts, supporting the investigation of the influence of anatomical variability and surgical planning on implant primary stability. The development of the computational framework was carried out in close collaboration with the University of Cagliari, whose clinical expertise contributed to the definition of clinically realistic surgical planning strategies and to the interpretation of the simulated outcomes. This collaboration ensured that the numerical analyses reflected conditions commonly encountered in routine orthopaedic practice and strengthened the translational relevance of the developed workflow. To ensure scalability and computational efficiency, the automated pipeline was successfully ported to high-performance computing infrastructures, in particular to the CINECA Leonardo supercomputer. The use of high-performance computing resources enabled the execution of subject-specific simulations in a time-effective manner and represented a key enabling factor for large-scale in silico studies within the IST4HR project. Overall, the activities carried out by UNIBO resulted in a robust, scalable and clinically informed computational framework for the assessment of implant primary stability, representing a fundamental component of the IST4HR in silico trial environment and providing a solid basis for retrospective validation and extended in silico investigations.
Dettagli del progetto
Responsabile scientifico: Marco Viceconti
Strutture Unibo coinvolte:
Dipartimento di Ingegneria Industriale
Coordinatore:
Università di PISA(Italy)
Contributo totale Unibo: Euro (EUR) 81.876,00
Durata del progetto in mesi: 24
Data di inizio
28/09/2023
Data di fine:
28/02/2026
