Bone is capable of adapting to external loads by reorganizing its architecture as well as mass. This metabolic process is called bone remodeling or adaptation. Since the 1970s, researchers in the field have been attempting to propose a mathematical model that can capture this complex process. Homeostasis-based phenomenological models have been widely used for bone remodeling simulation on the biological foundation. As an alternative, optimization-based simulations have been recently studied to express time-dependent bone remodeling phenomenon. In this paper, we proposed a novel optimization-based method for age-related bone remodeling by introducing two parameters in the optimization formulation. The proposed method was validated through a bone specimen in comparison with the relevant data in the literature. It was then applied to human proximal femur to simulate age-related changes in trabecular bone in the age span of 20 to 60 years. A quantitative comparison was made in terms of BV/TV. Other morphological indices, Tb.N and DA, were also in line with the data in the literature. Further in-depth studies would enhance the predictive capability of the proposed method and thus contribute to the development of patient-specific model.
Computational bone remodeling simulation has been widely studied and markedly evolved to predict trabecular architecture alteration under various load conditions. Subject-specific simulation, however, still remains to be further investigated. The bottleneck has been the determination of physiological loads for a subject-specific bone specimen. In the field, several inverse load estimation algorithms have been proposed to estimate “unknown” loads with the information on the “known” bone density distribution (or bone architecture). Shifting the current paradigm of load estimation, we may estimate local loads for the region of interest (ROI) with the given global loads. In the finite element method, a technique called ‘substructuring’ has been used to link between an original full model and separate substructures, and thus to handle a very large FE model with limited computing resources. In this paper, we proposed a novel method of estimating local loads for the ROI based on the concept of substructuring. Using a synthetic proximal femur for load estimation, the proposed method was validated in terms of force equilibrium, displacement field, and strain energy density (SED) distribution. Moreover, localized bone remodeling simulation with the estimated local loads successfully produced similar trabecular morphology compared with that for the ROI’s in the synthetic proximal femur. From the numerical results, it could be concluded that, with further in vivo load measurement research, the proposed method would contribute to subject-specific bone remodeling simulation.
Since Roentgen’s discovery of x-ray, medical imaging devices have evolved to scan three-dimensional internal organic structure. For timely diagnosis of osteoporosis, high-resolution trabecular scan is essential. However, radiologic imaging techniques inherently imposes higher radiation exposure when the desired resolution is high. Various image-processing techniques have been researched to overcome this problem and are being commercialized. Nevertheless, these techniques generally require additional scans for reliable image refinement. As an alternative approach, using bone remodeling simulation the trabecular architecture can be accurately predicted. In this research, using low-resolution density distribution as a constraint, trabecular image was reconstructed and the results were consistent with the original images. This framework can used to reconstruct QCT images so that the detailed microstructure would become visible, enabling clinicians diagnose osteoporosis more accurately.
전세계적 노령화와 함께 골다공증과 같은 퇴행성 골질환이 사회적 문제가 되고 있다. 이러한 퇴행성 골질환의 특징은 그 과정이 비가역적이라 어떠한 약물로도 완전한 회복이 어려다는 점이다. 이런 측면에서 치료보다는 예방의 중요성이 매우 강조되고 있다. 그러나 현재로선 골밀도만을 기준으로 골다공증 진단이 이루어져 조기진단이 매우 어려운 실정이다. 골강도의 정량적인 파악을 위해서는 골밀도 뿐 아니라 해면골의 미세구조 분석이 필요하다. 골재형성에 의한 미세구조 변화를 전산모사를 통해 예측할 수 있다면 골다공증 조기진단, 예후예측, 환자 맞춤 처방에 큰 기여를 할 것으로 기대된다.
본 연구에서는 골재형성 전산모사에 필수적인 세 가지 요소에 대한 연구를 수행하였다. 첫째로는 골재형성에 대한 수학적 모델이다. 제안된 방법론은 위상최적설계기반으로 기존 타 연구그룹에서 제안한 모델과는 수식화 측면에서 다르다. 그럼에도 본 연구에서는 제안된 모델이 타 모델과 동일한 기능을 가지는 것을 입증하였으며 대퇴골 해면골 미세구조의 골재형성 과정을 매우 정확히 모사하였다. 두 번째 요소는 하중조건이다. 골재형성이 국부 골조직에 대해서 이루어지고 있는 반면 국부 골조직 주변 하중조건은 측정이나 계산이 매우 어려웠다. 본 연구에서는 전체 하중조건을 안다는 가정하, 국부 하중조건을 계산할 수 있는 방법론을 제안하였다. 제안 방법론은 기존 방법론과 비교해 보다 정확하고 효율적임을 보였다. 마지막 세 번째 요소는 앞 두 요소를 이용할 골영상 고해상화로의 응용이다. 본 연구에서 저해상의 밀도분포, 국부하중, 골재형성 전산모사를 이용하여 해면골 영상의 고해상화가 가능함을 보였다. 이러한 방법론은 기존 영상처리기법을 이용한 고해상화와 크게 차별되어 영상의학계에 획기적인 기술이라 할 수 있다.
연구결과들로 미루어 보아, 본 연구가 향후 골재형성 전산모사 발전, 그리고 임상에의 적용에 기여할 수 있을 것으로 예상한다.
