Comparison of mathematical morphological descriptors of hyporeflective cavities in optical coherence tomography of patients with macular telangiectasia compared to patients with diabetic maculopathy

Editor, I maging with optical coherence tomography (OCT) in patients with macular telangiectasia type 2 (MacTel) demonstrates typical hyporeflective cavities of the inner and outer neurosensory retina. (Charbel Issa et al. 2013) It is proposed that these cavities result from cellular atrophy in the middle retinal layers (Gass & Blodi 1993), and clinical observation suggests that they have a different morphology to cavities in patients with exudative vascular pathologies. In particular, it has been noted that the hyporeflective cavities in MacTel are more oblong, less circular and more horizontally orientated than typical cysts in diabetic macular oedema (DME). (Ryan 2006) Distinguishing such morphological differences in OCT cavities using objective mathematical terms may provide insight into pathology and could potentially assist towards differentiating degenerative OCT spaces from those caused by leakage in various complex clinical cases, without the need for more invasive techniques. This differentiation has a profound impact on the likely management pathway for such cases (Fig. 1). Following sample size calculations, images from nine sequential patients with MacTel and 9 with DME were utilized for a comparative experimental study. A single horizontal OCT scan nearest to the centre of the fovea was exported for each eligible eye; if two eyes were eligible, only the right eye was recruited to the study. Images were anonymized and analysed by a second masked researcher using previously validated MatLab algorithms. (Aslam et al. 2009) An additional median filtering operation was performed to remove noise artefact, and morphological descriptors of delineated cavities were derived by the software, namely circularity, eccentricity and orientation. Circularity is defined as the ratio of perimeterof thearea specifieddividedby the perimeter if that same area were in the shape of a perfect circle. Eccentricity is the ratio of the distance between the foci of the ellipse and its major axis length. As dimensionless indices, these measures are resistant to systematic error from varying imaging modalities ormagnification. Statistical comparison of the morphological descriptors between groups was performed according to the clinical hypothesis with a onetail t-test, after assessing for normality (Shapiro–Wilks test). In accordance with sample size calculations, 16 cavities from nine eyes of patients with MacTel and 17 cavities from nine eyes of patients with DME were subjected to image analysis using MatLab algorithms. As hypothesized, the cavities in patientswithMacTelwere found tobe significantly less circular and more eccentric compared to those with DME (p = 0.038 for circularity and p = 0.019 for eccentricity). No significant difference was found in orientation of the cysts, although the study was not powered towards assessing this outcome measure. This study demonstrates a significant difference in mathematical shape characteristics of OCT hyporeflective spaces between patients with exudative cysts of DME and those with the presumed atrophic degenerative cysts of MacTel. The measurements were made using dimensionless indices, and this paper represents the first study to demonstrate significant differences in objective mathematical shape descriptors of OCT hyporeflective areas between differing pathologies. The findings add weight to the evidence suggesting that a separate aetiology is responsible for the formation of the cavities in each of the two clinical conditions studied. The measures described here could potentially be incorporated with other OCT image analysis descriptors (Barthelmes et al. 2008) for algorithms to help distinguish atrophic cavities from exudative cysts, without the need for invasive fluorescein angiography. These could be relevant for a variety of complex cases involving conditions such as toxic maculopathies, genetic disease and chronic macular degeneration.