Abstract
Nailfold capillaroscopy (NC) is a non-invasive imaging technique employed to assess the condition of blood capillaries in the nailfold. It is particularly useful for early detection of scleroderma spectrum disorders and evaluation of Raynaud's phenomenon. While automated approaches to analysing NC images are relatively rare, they are typically based on extraction and analysis of individual capillaries from the images in order to assign a patient to one of the commonly employed scleroderma patterns. In this chapter, we present a different approach that does not rely on individual capillaries but performs interpretation in a holistic way based on information gathered from an image or a selected image region. In particular, our algorithm employs texture analysis to characterise the underlying patterns, coupled with a classification stage to first identify patterns in fingers, and then, through a voting strategy, reach a decision for a patient. Experimental results on a set of NC images with known ground truth demonstrate the efficacy of the proposed approach.
TopNailfold Capillaroscopy
Capillaroscopy is an established technique to investigate micro-vascular involvement in various diseases. Examination of capillaries for finding a relation between conjunctival inflammation and the presence of an inextricable knot of capillary loops was noted by Italian physician Giovanni Rasori about 200 years ago using a magnifying glass (Cutolo et al., 2003). In 1911, Lombard discovered that human skin capillaries can be observed using a microscope after the application of a drop of immersion oil. Further to this, Weiss, in 1916, was able to take a picture of capillaries using a primordial camera. In 1925, Brown and O'Leary have shown the use of capillaroscopy for observing capillary abnormalities in Raynaud's phenomenon (RP) characterised by systemic sclerosis. Nevertheless, capillaroscopy was then mostly neglected for several decades until, in 1973, Maricq and LeRoy published the first paper describing specific capillaroscopic patterns in SSc (Maricq & LeRoy, 1973).
Key Terms in this Chapter
Support Vector Machine (SVM): Supervised learning model.
Nailfold Capillaroscopy: Non-invasive imaging technique employed to assess the condition and morphology of capillaries in the nailfold.
Local Binary Patterns (LBP): Texture features which encode the textural characteristics in the form of binary patterns.
Texture Features: Set of numerical values which encode textural characteristic of an image.
Texture: Visual or tactile surface characteristics and appearance of an object.
Raynaud’s Phenomenon: Condition characterised by reduction in blood flow in response to cold or stress.
Multi-Dimensional LBP: LBP features extracted at multiple scales and recorded into a multi-dimensional histogram.
Systemic Sclerosis: An autoimmune or connective tissue disease characterised by thickening of the skin.