Automated segmentation and analysis of normal and osteoarthritic knee menisci from magnetic resonance images – data from the Osteoarthritis Initiative

Paproki, Anthony; Engstrom, Craig; Chandra, Shekhar S.; Neubert, Aleš; Fripp, Jürgen; Crozier, Stuart

doi:10.1016/j.joca.2014.06.029

Cited by 38 publications

(17 citation statements)

References 41 publications

(47 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generation of a subject-specific computational model requires a lot of manual work and time in segmentation of soft tissues, meshing and making models to converge. In future studies, the methodology presented here should be coupled with semi-automatic or fully automatic segmentation techniques (Chandra et al, 2016;Dodin et al, 2010;Folkesson et al, 2007;Lee et al, 2014;Liukkonen et al, 2017b;Paproki et al, 2014;Shan et al, 2014;Tamez-Pena et al, 2012;Yang et al, 2015;Yu et al, 2016) and with automated meshing tools (Rodriguez-Vila et al, 2017). As motion capture systems are not readily available in clinical settings, a simple and fast method should be developed to obtain and implement patient's gait.…”

Section: Clinical Applicationmentioning

confidence: 99%

Identification of locations susceptible to osteoarthritis in patients with anterior cruciate ligament reconstruction: Combining knee joint computational modelling with follow-up T1ρ and T2 imaging

Bolcos

Mononen

Tanaka

et al. 2020

Clinical Biomechanics

View full text Add to dashboard Cite

Background: Finite element modelling can be used to evaluate altered loading conditions and failure locations in knee joint tissues. One limitation of this modelling approach has been experimental comparison. The aims of this proof-of-concept study were: 1) identify areas susceptible to osteoarthritis progression in anterior cruciate ligament reconstructed patients using finite element modelling; 2) compare the identified areas against changes in T 2 and T 1ρ values between 1-year and 3-year follow-up timepoints. Methods: Two patient-specific finite element models of knee joints with anterior cruciate ligament reconstruction were created. The knee geometry was based on clinical magnetic resonance imaging and joint loading was obtained via motion capture. We evaluated biomechanical parameters linked with cartilage degeneration and compared the identified risk areas against T 2 and T 1ρ maps. Findings: The risk areas identified by the finite element models matched the follow-up magnetic resonance imaging findings. For Patient 1, excessive values of maximum principal stresses and shear strains were observed in the posterior side of the lateral tibial and femoral cartilage. For Patient 2, high values of maximum principal stresses and shear strains of cartilage were observed in the posterior side of the medial joint compartment. For both patients, increased T 2 and T 1ρ values between the follow-up times were observed in the same areas. Interpretation: Finite element models with patient-specific geometries and motions and relatively simple material models of tissues were able to identify areas susceptible to post-traumatic knee osteoarthritis. We suggest that the methodology presented here may be applied in large cohort studies.

show abstract

Section: Clinical Applicationmentioning

confidence: 99%

Identification of locations susceptible to osteoarthritis in patients with anterior cruciate ligament reconstruction: Combining knee joint computational modelling with follow-up T1ρ and T2 imaging

Bolcos

Mononen

Tanaka

et al. 2020

Clinical Biomechanics

View full text Add to dashboard Cite

show abstract

“…3D surface models must be generated for all allografts. In the future, meniscus segmentation could probably be performed in a semiautomated or fully automated manner [45,46], which could reduce additional costs. Automatized comparison of the meniscus template of the healthy contralateral meniscus and all available allografts is already possible.…”

Section: Basic Prerequisites For 3d Sizingmentioning

confidence: 99%

Three-dimensional meniscus allograft sizing—a study of 280 healthy menisci

et al. 2020

View full text Add to dashboard Cite

Background: Inaccurate meniscus allograft size is still an important problem of the currently used sizing methods. The purpose of this study was to evaluate a new three-dimensional (3D) meniscus-sizing method to increase the accuracy of the selected allografts. Methods: 3D triangular surface models were generated from 280 menisci based on 50 bilateral and 40 unilateral knee joint magnetic resonance imaging (MRI) scans. These models served as an imaginary meniscus allograft tissue bank. Meniscus sizing and allograft selection was simulated for all 50 bilateral knee joints by (1) the closest mean surface distance (MeSD) (3D-MRI sizing with contralateral meniscus), (2) the smallest meniscal width/length difference in MRI (2D-MRI sizing with contralateral meniscus), and (3) conventional radiography as proposed by Pollard (2D-radiograph (RX) sizing with ipsilateral tibia plateau). 3D shape and meniscal width, length, and height were compared between the original meniscus and the selected meniscus using the three sizing methods. Results: Allograft selection by MeSD (3D MRI) was superior for all measurement parameters. In particular, the 3D shape was significantly improved (p < 0.001), while the mean differences in meniscal width, length, and height were only slightly better than the allograft selected by the other methods. Outliers were reduced by up to 55% (vs. 2D MRI) and 83% (vs. 2D RX) for the medial meniscus and 39% (vs. 2D MRI) and 56% (vs. 2D RX) for the lateral meniscus. Conclusion: 3D-MRI sizing by MeSD using the contralateral meniscus as a reconstruction template can significantly improve meniscus allograft selection. Sizing using conventional radiography should probably not be recommended. Trial registration: Kantonale Ethikkommission Zürich had given the approval for the study (BASEC-No. 2018-00856).

show abstract

“…There are a few publications present image processing algorithms dedicated to the knee joint, but directed to evaluate cartilage [15] on weight bearing parts. Semi-automated and automated segmentation techniques of menisci on high field system are already described in two works [16,17]. But none with a use of techniques proposed in our study.…”

Section: Introductionmentioning

confidence: 92%

Image processing in detection of knee joints injuries based on MRI images

Kohut¹,

Holak²,

Obuchowicz³

2017

J. vibroeng.

View full text Add to dashboard Cite

This paper presents image processing methods for visualization and classification of medial meniscus tears. The first method uses watershed with a threshold segmentation approach. The algorithm was tested on a number of images of the knee obtained with a use of the magnetic resonance imaging technique (MR). Images of the knee were collected from healthy subjects and patients with a clinically diagnosed meniscal pathology. Then, watershed technique was compared with other popular methods of image segmentation, i.e. simple thresholding and region growing. For this purpose, the execution speed and the efficiency of the methods were analyzed. Additionally, an automatic detection of the meniscus based on MRI of the knee joint was developed. The solutions were implemented using classical image processing methods in the MATLAB environment with an application of the Image Processing Toolbox and MVtec Halcon vision libraries. The presented methods will have a practical value for the referring physicians and the diagnostic imaging specialists.

show abstract

Automated segmentation and analysis of normal and osteoarthritic knee menisci from magnetic resonance images – data from the Osteoarthritis Initiative

Cited by 38 publications

References 41 publications

Identification of locations susceptible to osteoarthritis in patients with anterior cruciate ligament reconstruction: Combining knee joint computational modelling with follow-up T1ρ and T2 imaging

Identification of locations susceptible to osteoarthritis in patients with anterior cruciate ligament reconstruction: Combining knee joint computational modelling with follow-up T1ρ and T2 imaging

Three-dimensional meniscus allograft sizing—a study of 280 healthy menisci

Image processing in detection of knee joints injuries based on MRI images

Contact Info

Product

Resources

About