User menu

Accès à distance ? S'identifier sur le proxy UCLouvain

Clinical evaluation of automated scan prescription of knee MR images.

Primary tabs

Lecouvet, Frédéric [UCL] Claus, Jérôme Schmitz, Patrick Denolin, Vincent Bos, Clemens Vande Berg, Bruno [UCL]

PURPOSE: To compare an automated scan planning method to manual scan positioning in routine knee magnetic resonance imaging (MRI) studies. MATERIALS AND METHODS: The automated scan planning method uses anatomical landmarks in a 3D survey of the knee. The method is trained by example plannings, consisting of manual slice positioning by an experienced technologist in 15 MRI studies. Automated knee MR examinations obtained in three geometries in 50 consecutive patients were compared to those obtained in 50 consecutive control patients, where imaging planes were planned manually. Anatomical coverage and slice angulation were scored for each geometry on a 4-grade scale by an experienced radiologist blinded to the way of planning; groups were compared using a Mann-Whitney U-test. RESULTS: In 150 automated sequences the technologist adapted slice positioning in four cases (addition of slices to adapt to the size of the knee), representing the only automated sequences that received a poor rating. Thirteen sequences with manual planning received a poor rating. No difference in quality was found (P > 0.05) between automated and manual plannings for coronal coverage, sagittal coverage and angulation, and transverse angulation. Rating of automated planning was higher for transverse coverage, but lower than manual planning for coronal angulation. CONCLUSION: Automated sequence prescription for knee MRI is feasible in clinical practice, with similar quality as manual positioning.

Document type Article de périodique (Journal article) – Evaluation Studies, Journal Article
Access type Accès restreint
Publication date 2009
Journal information "Journal of magnetic resonance imaging : JMRI" - Vol. 29, no. 1, p. 141-5 (2009)
Peer reviewed yes
issn 1053-1807
Publication status Publié
Affiliation UCL - MD/RAIM - Département de radiologie et d'imagerie médicale
MESH Subject Algorithms ; Artificial Intelligence ; Humans ; Image Enhancement - methods ; Image Interpretation, Computer-Assisted - methods ; Imaging, Three-Dimensional - methods ; Knee Joint - anatomy & histology ; Magnetic Resonance Imaging - methods ; Observer Variation ; Pattern Recognition, Automated - methods ; Reproducibility of Results ; Sensitivity and Specificity
Links
  1. Itti, Magn Reson Med, 45, 486 (2001)
  2. Young S, Bystrov D, Netsch T, et al. Robust anatomy recognition for automated MR neuro scan planning. In: Proc ISMRM, 14th Scientific Meeting, Seattle, 2006; p 1588.
  3. Benner, Radiology, 239, 246 (2006)
  4. Weiss, Radiology, 239, 255 (2006)
  5. Lelieveldt, Radiology, 221, 537 (2001)
  6. Bystrov B, Pekar V, Young S, Dries SPM, Heese HS, Van Muiswinkel AM. Automated planning of MRI scans of knee joints. In: Reinhardt JM, editor. Medical imaging: image processing. Proc SPIE Medical Imaging, 2006; p 6144-6158.
  7. Mink J H, Levy T, Crues J V, Tears of the anterior cruciate ligament and menisci of the knee: MR imaging evaluation., 10.1148/radiology.167.3.3363138
  8. Buckwalter K A, Pennes D R, Anterior cruciate ligament: oblique sagittal MR imaging., 10.1148/radiology.175.1.2315495
  9. Stoller, Magnetic resonance imaging in orthopaedics and sports medicine, 305 (2006)
  10. van der Kouwe, Neuroimage, 27, 222 (2005)
Bibliographic reference Lecouvet, Frédéric ; Claus, Jérôme ; Schmitz, Patrick ; Denolin, Vincent ; Bos, Clemens ; et. al. Clinical evaluation of automated scan prescription of knee MR images.. In: Journal of magnetic resonance imaging : JMRI, Vol. 29, no. 1, p. 141-5 (2009)
Permanent URL http://hdl.handle.net/2078.1/25736