TY - JOUR AU - Carichino, Lucia AU - Guidoboni, Giovanna AU - Verticchio Vercellin, Alice Chandra AU - Milano, Giovanni AU - Cutolo, Carlo Alberto AU - Tinelli, Carmine AU - De Silvestri, Annalisa AU - Lapin, Sergey AU - Gross, Joshua C. AU - Siesky, Brent A. AU - Harris, Alon PY - 2016/12/15 Y2 - 2024/03/29 TI - Computer-aided identification of novel ophthalmic artery waveform parameters in healthy subjects and glaucoma patients JF - Modeling and Artificial Intelligence in Ophthalmology JA - MAIO VL - 1 IS - 2 SE - Original Articles DO - 10.35119/maio.v1i2.31 UR - https://www.maio-journal.com/index.php/MAIO/article/view/31 SP - 59-69 AB - <div class="page" title="Page 1"><div class="layoutArea"><div class="column"><p><em>Purpose:</em> Arterial waveform parameters (WPs) are commonly used to monitor and diagnose systemic diseases. Color Doppler Imaging (CDI) is a consolidated technique to measure blood velocity profile in some of the major ocular vessels. This study proposes a computer-aided manipulation process of ophthalmic artery (OA) CDI images to classify and quantify WPs that might be significant in the assessment of glaucoma.</p><p><em>Methods:</em> Fifty CDI images acquired by four different operators on nine healthy individuals and 38 CDI images of 38 open-angle glaucoma (OAG) patients were considered. An ad-hoc semi-automated image processing code was implemented to detect the digitalized OA velocity waveform and to extract the WPs. Concordance correlation coefficient (CCC), two-sample t-test and Pearson’s correlation coefficient were used to test for similarities, differences and associations among variables.</p><p><em>Results:</em> The OA-CDI images manipulation proposed showed a higher concordance between measured peak systolic velocity (PSV) data and extracted PSV data (0.80≤CCC≤0.98) than on end diastolic velocity (EDV) (0.45≤CCC≤0.63) and resistive index (RI) (0.30≤CCC≤0.58) data. In OAG patients, EDV, RI, subendocardial viability ratio (SEVR), period (T), area ratio (f) and normalized distance between ascending and descending limb (DAD/T) were found statistically correlated to at least one of the following factors: gender, age, ocular medications and year of diagnosis. When compared to healthy individuals, OAG patients OA-CDI profiles showed statistically higher values of f (p &lt; 0.001) and DAD/T (p = 0.002) (p-values corrected by age and gender).</p><p><em>Conclusion:</em> The proposed computer-aided manipulation of OA-CDI images allowed to identify DAD/T as a novel WP that vary significantly among healthy individuals and OAG patients, and among female and male OAG patients. Future studies on longitudinal OAG data are suggested to investigate the potential of DAD/T to predict severity and progression of the disease.</p></div></div></div> ER -