<p>In medical image analysis, the cost of acquiring high-quality data and their annotation by experts is a barrier in many medical applications. Most of the techniques used are based on supervised learning framework and need a large amount of annotated data to achieve satisfactory performance. As an alternative, in this paper, we propose a self-supervised learning (SSL) approach to learn the spatial anatomical representations from the frames of magnetic resonance (MR) video clips for the diagnosis of knee medical conditions. The pretext model learns meaningful spatial context-invariant representations. The downstream task in our paper is a class imbalanced multi-label classification. Different experiments show that the features learnt by the pretext model provide competitive performance in the downstream task. Moreover, the efficiency and reliability of the proposed pretext model in learning representations of minority classes without applying any strategy towards imbalance in the dataset can be seen from the results. To the best of our knowledge, this work is the first work of its kind in showing the effectiveness and reliability of self-supervised learning algorithms in class imbalanced multi-label classification tasks on MR videos.</p>

No abstract

: This paper presents an image feature representation using geometrical feature descriptor for image recognition. Geometrical feature representation is primarily used in shape descriptors feature representation. The shape feature representation using Bezier curve representation and control point selection is defined. Curve geometry is used in feature representation defining the curvature details of an image. A new linear Bezier curve control point feature is proposed to improve the control point detection for feature representation. The accuracy of control point extraction is made by a linear threshold approach using Bezier curve. The performance w. r. t. retrieval accuracy and feature overhead outlines the significance of the proposed approach in geometrical feature representation and image retrieval.

No abstract

No abstract