2016/02/10 by Nevine Demitri, Abdelhak M. Zoubir, Demitri, Nevine +1
Biochemistry, Genetics and Molecular Biology · Medicine · #Applications (stat.AP) #Diabetes Management and Research #FOS: Computer and information sciences #Optical Imaging and Spectroscopy Techniques #Spectroscopy Techniques in Biomedical and Chemical Research
paper · pdf · doi:10.48550/arxiv.1602.03386
openalex publication_date 2016/02/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Glucometers present an important self-monitoring tool for diabetes patients\nand therefore must exhibit high accu- racy as well as good usability features.\nBased on an invasive, photometric measurement principle that drastically\nreduces the volume of the blood sample needed from the patient, we present a\nframework that is capable of dealing with small blood samples, while\nmaintaining the required accuracy. The framework consists of two major parts:\n1) image segmentation; and 2) convergence detection. Step 1) is based on\niterative mode-seeking methods to estimate the intensity value of the region of\ninterest. We present several variations of these methods and give theoretical\nproofs of their convergence. Our approach is able to deal with changes in the\nnumber and position of clusters without any prior knowledge. Furthermore, we\npropose a method based on sparse approximation to decrease the computational\nload, while maintaining accuracy. Step 2) is achieved by employing temporal\ntracking and prediction, herewith decreasing the measurement time, and, thus,\nimproving usability. Our framework is validated on several real data sets with\ndifferent characteristics. We show that we are able to estimate the underlying\nglucose concentration from much smaller blood samples than is currently\nstate-of-the- art with sufficient accuracy according to the most recent ISO\nstandards and reduce measurement time significantly compared to\nstate-of-the-art methods.\n