The use of active shape models for making thickness measurements of articular cartilage from MR images

Solloway, Stuart; Hutchinson, Charles E.; Waterton, John C.; Taylor, Chris

doi:10.1002/mrm.1910370620

Cited by 130 publications

(72 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They combine the segmentation technique with three-dimensional (3-D) image registration to detect changes in cartilage volume [14]. Solloway et al [15] use active shape models for slice-by-slice cartilage segmentation, and estimate cartilage thickness in the direction perpendicular to the medial axis in each slice.…”

Section: A Related Workmentioning

confidence: 99%

Segmenting Articular Cartilage Automatically Using a Voxel Classification Approach

Folkesson

Dam

Olsen

et al. 2007

IEEE Trans. Med. Imaging

167

159

View full text Add to dashboard Cite

Abstract-We present a fully automatic method for articular cartilage segmentation from magnetic resonance imaging (MRI) which we use as the foundation of a quantitative cartilage assessment. We evaluate our method by comparisons to manual segmentations by a radiologist and by examining the interscan reproducibility of the volume and area estimates. Training and evaluation of the method is performed on a data set consisting of 139 scans of knees with a status ranging from healthy to severely osteoarthritic. This is, to our knowledge, the only fully automatic cartilage segmentation method that has good agreement with manual segmentations, an interscan reproducibility as good as that of a human expert, and enables the separation between healthy and osteoarthritic populations. While high-field scanners offer high-quality imaging from which the articular cartilage have been evaluated extensively using manual and automated image analysis techniques, low-field scanners on the other hand produce lower quality images but to a fraction of the cost of their high-field counterpart. For low-field MRI, there is no well-established accuracy validation for quantitative cartilage estimates, but we show that differences between healthy and osteoarthritic populations are statistically significant using our cartilage volume and surface area estimates, which suggests that low-field MRI analysis can become a useful, affordable tool in clinical studies.

show abstract

Section: A Related Workmentioning

confidence: 99%

Segmenting Articular Cartilage Automatically Using a Voxel Classification Approach

Folkesson

Dam

Olsen

et al. 2007

IEEE Trans. Med. Imaging

167

159

View full text Add to dashboard Cite

show abstract

“…Fully manual surface extraction and segmentation 18,21,[29][30][31][32] is tedious, time consuming, and prone to subjective judgment. Fully automated processes 36,37 at present cannot accurately and reproducibly extract and segment articular cartilage layers in areas where the boundary is not sufficiently distinct or in noisy images.…”

Section: Introductionmentioning

confidence: 99%

Quantitative and topographical evaluation of ankle articular cartilage using high resolution MRI

Millington

Tang

et al. 2006

Journal Orthopaedic Research

View full text Add to dashboard Cite

The objectives of this study were to quantitatively evaluate the articular cartilage layers of the ankle and describe the cartilage topographical distribution across the joint surfaces using high resolution MRI and image segmentation. An anisotropic diffusion noise reduction algorithm and a directional gradient vector flow (dGVF) snake segmentation algorithm were applied to cartilage sensitive MR images. Eight cadaveric ankles were studied. Six repeated data sets were acquired in five of the ankles. Quantitative parameters were calculated for each cartilage layer; coefficients of variation (CV) were calculated from the six repeated data sets; and 3D thickness distribution maps were generated. The noise reduction algorithm produced marked image enhancement. Mean cartilage thickness ranged from 0.91 AE 0.08 mm in the fibula to 1.34 AE 0.14 mm in the talus. Mean cartilage volume was 3.32 AE 0.55 ml, 1.72 AE 0.25 ml, and 0.35 AE 0.06 ml for the talus, tibia, and fibula, respectively. Mean CV ranged 2.82%-5.04% for quantitative parameters in the talus and tibia. The reported noise reduction and segmentation technique allow precise extraction of ankle cartilage and 3D reconstructions show that the thickest cartilage occurs over the talar shoulders, where osteochondritits dissecans (OCD) lesions commonly occur. ß

show abstract

“…In the knee, which displays cartilage thickness values of 5 mm and more, it has been shown that the cartilage volume can be determined with a high degree of accuracy [9,11,12,27,30,33]. Specific 3D digital image analysis techniques have also permitted to compute the cartilage thickness and its variation throughout knee joint cartilage plates [5,38,39]. These analyses have been validated in comparison with anatomical sectioning [9,37], CT arthrography [I 1,121, A-mode ultrasound [S], and stereophotogrammetry [5].…”

mentioning

confidence: 99%

Quantitative cartilage imaging of the human hind foot: Precision and inter‐subject variability

Al-Ali

Graichen

Faber

et al. 2002

Journal Orthopaedic Research

View full text Add to dashboard Cite

Alterations of ankle cartilage are observed in degenerative and inflammatory joint disease, but cartilage cannot be directly visualized by radiography. The purpose of this study was therefore to analyze the feasibility and precision of quantitative cartilage imaging in the human hind foot (talocrural, talotarsal, and intertarsal joints), and to report the inter-subject variability for cartilage volume, thickness and surface areas. The feet of 16 healthy volunteers were imaged using a 3D gradient-echo magnetic resonance imaging sequence with water-excitation. After interpolation to a resolution of 1 z 0.125 y 0.125 mm? the cartilage plates were segmented, and the cartilage volume, thickness, and surface areas determined. The precision (four repeated measurements) was examined in eight volunteers, the RMS average CV% being 2.1% to 10.9% in single joint surfaces, and < 3% for the cumulative values of all joints. The mean cartilage thickness ranged from 0.57 i 0.08 (navicular surface) to 0.89 5 0 . 1 9 mm (trochlear surface for tibia). In conclusion this study shows that it is feasibIe to quantify thin cartilage layers in the hind foot under in vivo imaging conditions, and that the precision errors are substantially smaller than the inter-subject variability in healthy subjects.

show abstract

The use of active shape models for making thickness measurements of articular cartilage from MR images

Cited by 130 publications

References 31 publications

Segmenting Articular Cartilage Automatically Using a Voxel Classification Approach

Segmenting Articular Cartilage Automatically Using a Voxel Classification Approach

Quantitative and topographical evaluation of ankle articular cartilage using high resolution MRI

Quantitative cartilage imaging of the human hind foot: Precision and inter‐subject variability

Contact Info

Product

Resources

About