A Novel Approach to Classify Nailfold Capillary Images in Indian Population Using USB Digital Microscope

A Novel Approach to Classify Nailfold Capillary Images in Indian Population Using USB Digital Microscope

Suma K. V. (Department of Electronics and Communication, Ramaiah Institute of Technology, Bangalore, India), Vishwajit Sasi (Department of Electronics and Communication, Ramaiah Institute of Technology, Bangalore, India) and Bheemsain Rao (Crucible of Research and Innovation, PES University, Bangalore, India)
Copyright: © 2018 |Pages: 15
DOI: 10.4018/IJBCE.2018010102
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Nailfold capillaroscopy is a simple, non-invasive clinical aid used for the diagnosis of vascular dysfunction. The method proposed emphasizes on the USB Digital Microscope which is a cheaper and useful alternative to commercially available expensive videocapillaroscopes which produces high quality images. As the process of identifying anomalies in nailfold capillaries is a tedious and time-consuming process this article proposes a completely automated system to diagnose anomalies. The nailfold capillary images are pre-processed to highlight the important features and remove noise in the images. The processed images are then used to train machine learning models. This article then assays and compares the performance of Logistic Regression classifier, fully connected neural network, Convolutional Neural Network (CNN) and Random Forest classifiers by evaluating their classification accuracy, sensitivity and specificity. The results prove Logistic Regression to be most accurate with a low classification error rate of 10.64%. while, a substantial classification accuracy of 72% was obtained with a small dataset by using bottleneck features of a deep CNN.
Article Preview

Introduction

Capillaries are the sites in the circulatory system where the actual exchange of nutrients and wastes along with gases O2 & CO2 take place. They are ultra-thin blood vessels and can be imaged conveniently at the nail bed region of fingers since they appear parallel to the skin at the nailfold. They appear as a bunch of inverted-U shaped and uniformly spaced blood vessels in healthy individuals. In diseased subjects, the capillaroscopic pattern could consist of loss of capillaries and other physical anomalies. Specific capillaroscopic patterns are identified in various diseases. Hypertensive subjects seem to have low capillary density and presence of elongated capillaries while in Diabetes there is rampant occurrence of capillary disorganizations such as giant capillaries, bushy capillaries and avascular regions. In Scleroderma, three stages of severity namely ‘Early’, ‘Active’, and ‘Late’ are identified by Cutolo et al (Cutolo, 2013).

Nailfold capillaroscopy is a simple, non-invasive, inexpensive and repeatable clinical test. It has immense potential as a clinical tool for diagnosis, prognosis, research and therapeutic use in varied spectrum of diseases (Eduardo, 2015). The nailfold capillary (NFC) images are acquired using the microscope with a magnification of 20x onwards, but the capillaries are visible better when the magnification is 200x. Multiple types of equipment are used for this purpose. Reuven Bergman et al (2003) have used the unmodified hand-held dermatoscope as a capillaroscopic instrument to detect a scleroderma - dermatomyositis pattern. They found that the capillaroscopic results obtained with the dermatoscope can help the dermatologist in the clinical diagnosis of connective tissue disorders. Stereomicroscope has been used as a method of nailfold capillaroscopy (NFC) which is applicable to a wide age range of paediatric patients. H J Anders et al (2001) have conducted a prospective study comparing nailfold capillaroscopy using an ophthalmoscope and stereomicroscope for differentiating between primary and secondary Raynaud’s phenomenon. Nailfold videocapillaroscopy (NVC), is the current gold standard for detection of capillary abnormalities which produce superior quality images. NVC is expensive and hence in developing countries like India an inexpensive, yet reliable equipment is in need. This is especially relevant given the inclusion of capillaroscopic abnormality within the 2013 classification criteria for SSc (Van Den, 2013). Vivek Vasdev et al (2011) have proposed the use of USB Digital Microscope for acquiring NFC images with high reliability.

The novelty of our work lies in two aspects: one, in the use of nailfold capillary images rather than the video sequence as in nailfold videocapillaroscope and two, in the automation methodology employed for classifying these images based on the health condition. We propose an efficient computer automated diagnosing system using the images obtained from the USB digital microscope as depicted in Figure 1.

Figure 1.

Classification system

Motivation

Nailfold videocapillaroscope which is used for acquiring images is expensive and requires trained staff to interpret the capillary images for diagnosis of the health condition of the subject. We address these two aspects by acquiring the capillary images from an economical device called USB digital microscope and building an automated classification system which can avoid the need for trained staff to interpret the images. This classification system is a useful clinical tool which can be used for purposes such as diagnosis, prognosis, therapeutic study and research.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 8: 2 Issues (2019): Forthcoming, Available for Pre-Order
Volume 7: 2 Issues (2018): 1 Released, 1 Forthcoming
Volume 6: 2 Issues (2017)
Volume 5: 2 Issues (2016)
Volume 4: 2 Issues (2015)
Volume 3: 2 Issues (2014)
Volume 2: 2 Issues (2013)
Volume 1: 2 Issues (2012)
View Complete Journal Contents Listing