Automated Analysis of Cartilage Thickness and Tissue Quality from MRI

骨性关节炎是导致残疾和生活质量下降的主要原因。它与关节(尤其是膝盖)的骨骼和软骨的退行性变化有关，导致疼痛和残疾。这种疾病的变异性极大，一些患者病情恶化得非常快，而另一些患者则在多年后进展缓慢。目前还没有针对这种疾病的特别有效的治疗方法，但更有效的治疗方法即将出现--特别是针对软骨或骨骼退化的药物。疾病进展的临床指标，如疼痛和残疾，是主观的和非特异性的，导致漫长而昂贵的临床试验，最终可能被证明是不确定的。因此，迫切需要定量的生物标记物(疾病进展的衡量标准)来支持OA药物开发的决策。治疗骨性关节炎的有效但昂贵的药物的可获得性也将产生对确定哪些患者将从治疗中受益最大的方法的需要，从而导致更具成本效益的部署。该项目的目标是开发技术，通过对膝关节3D磁共振图像(MRI)的自动分析，为OA提供生物标志物。这将建立在该财团成员以前进行的可行性工作的基础上。具体目标是：1.从膝关节的核磁共振自动提取骨和软骨的结构--实现软骨和骨骼形态的计算机化测量；2.开发从图像测量软骨和骨质量的方法--能够及早发现退行性疾病；3.调整现有方法，以获得解剖学上一致的软骨厚度和质量图--能够对不同对象的图进行直接比较和统计分析；4.开发统计方法，检测这些图中微妙但一致的变化模式--提供对疾病过程的洞察和检测患者队列进展的灵敏手段；5.整合各种方法以创建技术演示器--为技术提供试验台；6.使用来自大规模办公自动化成像研究的数据验证该方法。

Osteoarthritis (OA) is a major cause of disability and reduced quality of life. It is associated with degenerative changes in the bone and cartilage of articulating joints (particularly the knee), leading to pain and disability. The disease is extremely variable, with some patients deteriorating extremely rapidly, whilst others progress slowly over many years. There are currently no particularly effective treatments for the disease, but more effective therapies are on the horizon - particularly drugs that target cartilage or bone degeneration. Clinical indicators of disease progression, such as pain and disability, are subjective and non-specific, leading to long and expensive clinical trials, which may ultimately prove inconclusive. There is thus a pressing need for quantitative biomarkers (measures of disease progression) to support decision making in OA drug development. The availability of effective but expensive drugs for OA will also create a need for methods to identify those patients who would most benefit from treatment, leading to more cost-effective deployment. The aim of the project is to develop technology to provide biomarkers for OA, through automated analysis of 3D magnetic resonance images (MRI) of the knee. This will build on feasibility work previously undertaken by members of the consortium. The specific objectives are to: 1. Extract the structure of bone and cartilage from MRI of the knee automatically - enabling computerised measurement of cartilage and bone morphology; 2. Develop methods for measuring cartilage and bone quality from the images - enabling early detection of degenerative disease; 3. Adapt existing methods for obtaining anatomically consistent maps of cartilage thickness and quality - enabling direct comparison and statistical analysis of maps from different subjects; 4. Develop statistical methods for detecting subtle but consistent patterns of change in these maps - providing insight into the disease process and a sensitive means of detecting progression in cohorts of patients; 5. Integrate the methods to create a technology demonstrator - providing a test-bed for the technology; 6. Validate the approach using data from large-scale OA imaging studies.