Unified hypergraph for image ranking in a multimodal context

SummaryBackground: Magnetic resonance (MR) images of the postoperative canine stifle are adversely affected by susceptibility artefacts associated with metallic implants.Objectives: To determine empirically to what extent susceptibility artefacts could be reduced by modifications to MR technique.Methods: Three cadaveric limbs with a tibial plateau levelling osteotomy (TPLO), tibial tuberosity advancement (TTA), or extra-capsular stabilization (ECS) implant, respectively, were imaged at 1.5T. Series of proton density and T2-weighted images were acquired with different combinations of frequency-encoding gradient (FEG) direction and polarity, stifle flexion or extension, echo spacing (ES), and readout bandwidth (ROBW), and ranked. The highest rank (a rank of 1) corresponded to the smallest artefact.Results: Image ranking was affected by FEG polarity (p = 0.005), stifle flexion (p = 0.01), and ROBW (p = 0.0001). For TPLO and TTA implants, the highest ranked images were obtained with the stifle flexed, lateromedial FEG, and medial polarity for dorsal images, and craniocaudal FEG and caudal polarity for sagittal images. For the ECS implant, the highest ranked images were obtained with the stifle extended, a proximodistal FEG and proximal polarity for dorsal images, and craniocaudal FEG and cranial polarity for sagittal images.Clinical significance: Susceptibility artefacts in MR images of postoperative canine stifles do not preclude clinical evaluation of joints with ECS or TTA implants.Part of this study was presented at the Annual Meeting of the American College of Veterinary Radiology, Albuquerque, NM, October 2011.

Download Full-text

Enhancing Visual Place Inference via Image Ranking Using Fuzzy Similarity and Agreeability

10.1007/978-3-030-86271-8_34 ◽

2021 ◽

pp. 403-414

Author(s):

Piotr Wozniak ◽

Bogdan Kwolek

Keyword(s):

Fuzzy Similarity ◽

Image Ranking

Download Full-text

Ordinal Distance Metric Learning with MDS for Image Ranking

Asia Pacific Journal of Operational Research ◽

10.1142/s0217595918500070 ◽

2018 ◽

Vol 35 (01) ◽

pp. 1850007 ◽

Cited By ~ 1

Author(s):

Panpan Yu ◽

Qingna Li

Keyword(s):

Data Structure ◽

Metric Learning ◽

Learning Model ◽

Distance Metric Learning ◽

Distance Metric ◽

Local Data ◽

Linear Distance ◽

Image Ranking ◽

Unconstrained Problem ◽

The Cost

Image ranking is to rank images based on some known ranked images. In this paper, we propose an improved linear ordinal distance metric learning approach based on the linear distance metric learning model. By decomposing the distance metric [Formula: see text] as [Formula: see text], the problem can be cast as looking for a linear map between two sets of points in different spaces, meanwhile maintaining some data structures. The ordinal relation of the labels can be maintained via classical multidimensional scaling, a popular tool for dimension reduction in statistics. A least squares fitting term is then introduced to the cost function, which can also maintain the local data structure. The resulting model is an unconstrained problem, and can better fit the data structure. Extensive numerical results demonstrate the improvement of the new approach over the linear distance metric learning model both in speed and ranking performance.

Download Full-text