Interactive Journal of Medical Research
A general medical journal, focusing on innovation in health and medical research
i-JMR is a general medical journal with a focus on innovation in health, health care, and medicine - through new medical techniques and innovative ideas and/or research, including - but not limited to - technology, clinical informatics, or groundbreaking research.
Published by JMIR Publications, publisher of JMIR, the leading eHealth/mHealth journal (Impact Factor 2015: 4.532), i-JMR is a JMIR "sister journal" which features a rapid and thorough peer-review process, professional copyediting, professional production of PDF, XHTML, and XML proofs.
i-JMR is indexed in PubMed and archived in PubMed Central.
Jan 10, 2017
Oct 21, 2016
Aug 29, 2016
Aug 25, 2016
Jul 21, 2016
Jul 13, 2016
Jun 24, 2016
Jun 20, 2016
Jun 17, 2016
Jun 7, 2016
May 25, 2016
May 24, 2016
Citing this Article
Right click to copy or hit: ctrl+c (cmd+c on mac)
Latest Submissions Open for Peer-Review:View All Open Peer Review Articles
sickleCAPTCHA - Design and Testing a Touch-based CAPTHCA for Monitoring and Analysis of Sickle Cell Anemia: A Comparative Study
Date Submitted: Dec 13, 2016
Open Peer Review Period: Dec 14, 2016 - Feb 8, 2017
Background: Patients with Sickle Cell Anemia are monitored by scoring red blood cells according to their morphology in blood samples. Several image-processing algorithm for automated erythrocyte count...
Background: Patients with Sickle Cell Anemia are monitored by scoring red blood cells according to their morphology in blood samples. Several image-processing algorithm for automated erythrocyte counts and morphological analysis in blood samples has been developed. However, overlapping, occluded, or grouped erythrocytes can generate erroneous results. We designed and tested an approach that users identify red blood cells with specific morphologies in the images through a touch-based CAPTCHA (sickleCAPTCHA). Objective: To answer three research questions: (1) Does sickleCAPTCHA allow the accurate analysis of red blood cells from sickle cell anemia patients (reliability)?, (2) is sickleCAPTCHA a valid system of access control (security)?, and (3) is sickleCAPTCHA appropriate for touch devices (adequacy)? Methods: We designed touch-based CAPTCHA to locate, count, and identify red blood cells in a microscopy blood sample image. We tested sickeCAPTCHA with 101 university students. We then compared the responses of participants to the sickleCAPTCHA tasks with expert classifications of images from erythrocytesIDB database (reliability). We also compared the results of applying an image-processing algorithm with the expert classifications, to determine to what extent verified tasks could be resolved automatically (security). We then conducted a usability test, comparing with reCAPTCHA, to determine the adequacy on touch devices. Results: With respect to reliability, participants correctly analyze the morphology of 100% clusters of red blood cells contained in erythrocytesIDB. About security, image-processing algorithm only classified correctly 0.7% of the clusters. Regarding adequacy, results show that resolving a sickleCAPTCHA, on average, is faster than reCAPTCHA (6.95 ± 3.92, P<.001, on touch devices). We also performed an ad-hoc power test to calculate the minimum sample size required (90 participants). Conclusions: The results confirmed that sickeCAPTCHA is suitable for use on touch devices, the resulting morphological analysis is reliable, and access control is safe. As future work, we plan to study whether this system is suitable for the analysis of other types of morphologies and to apply sickleCAPTCHA in a real environment with a large number of daily users.