A quantitative study on the effect of applied probe pressure during OCTA skin acquisitions

Optical Coherence Tomography Angiography (OCTA) is an emerging non-invasive, high-resolution imaging technique used in dermatology for visualizing skin vasculature. Images are acquired through a probe brought in contact with the skin surface; however, this modality can introduce mechanical stress to the skin and temporarily change the underlying vascular structure, compromising the accuracy of OCTA images. Therefore, the pressure applied by the OCTA probe is a critical factor that can affect the repeatability of acquisitions and reliability of the resulting data. This thesis aims to analyze how OCTA probe pressure may affect the accuracy and quality of angiographic images, evaluating the impact via specific metrics and validating the results through statistical analysis. For this study, 10 acquisitions were taken on healthy volunteers at the Medical University of Vienna, in three different anatomical regions (arm, head, leg), for a total of 53 OCTA volumes. Data was taken by maintaining constant contact between probe and skin surface, but progressively increasing the applied force, in a range from 1 to 10 Newtons. The OCTA probe was equipped with a force sensor for real-time measurement and data collection during acquisition. Using data processing algorithms, OCTA images were reconstructed from the raw OCT volumes through an intensity-based-method, followed by a dedicated artifact attenuation pipeline. At this processing stage, the Mean Noise Level and PSNR (Peak Signal-to-Noise Ratio) were calculated to assess the amount of noise in the volumes. Then, from the OCTA volumes, two 2D projections were derived, one for the superficial layer and one for the deep layer. Additional image quality metrics (BRISQUE, NIQE and PIQE), texture parameters such as entropy and skewness, and vascular features such as vascular density were estimated. To evaluate the influence of pressure on image quality and to determine an ideal range pressure for each imaging location, statistical analyses, including linear regression and ANOVA test, were performed to assess differences between layers and anatomical regions. An increase in pressure is observed to have a greater impact on the superficial layer than on the deep layer. As the applied pressure increases, there is a noticeable change in noise-related parameters (pvalue<0.05), which appears to influence image characteristics. In contrast, a comparison between different acquisitions reveals a decrease in vascular density (around -2% arm, around -2% head, around -5% leg), suggesting a modification in structural details and highlighting how pressure variations can affect vascular features. Among the body locations analyzed, the head appears to be more sensitive to pressure, distinguished by greater variation in metrics. Ultimately, the optimal pressure range varies according to the anatomical region considered. However, the presence of artifacts in OCTA images due to involuntary movements is a limitation of this study, as is the reduced sample size analyzed. In addition, the analysis of only three body areas does not allow the results to be generalized to other skin regions. This thesis aims to establish guidelines for controlled pressure in OCTA acquisitions, enhancing image quality and parameter reliability for accurate analysis of superficial skin microvasculature. The results show that pressure noticeably impacts image quality and vascular features, with optimal ranges varying by anatomical region, providing a reference for standardized imaging protocols