A Landmark-Based Approach for Robust Estimation of Exposure Index Values in Digital Radiography

Abstract

International audienceThe exposure index (EI) gives a feedback to radiographers on the image quality in digital radiography, but its estimation on clinical images raises many challenges. In this paper we provide a critical overview of state of the art methods that address this problem and we show that more robust results can be obtained by detecting anatomical structures. This new approach implicitly manages the presence of multiple structures in the field-of-view. Moreover, we propose a landmark-based method that, by exploiting redundancy of local estimates, is more robust to potential detection errors

A Landmark Detection Approach Applied to Robust Estimation of the Exposure Index in Digital Radiography

“…if the estimated EI value increases by +25% and decreases by 20% with respect to the target EI value. Irrera et al (2015) have proposed to replace the ROI definition with a landmark based approach by associating a set of anatomical landmarks with struc-tures of interest. We summarize here how EI values are computed from these sets of points.…”

“…In (Irrera et al, 2015) landmarks were positioned manually and errors were simulated in order to show the interest of the method. In this paper, we propose a technique to detect these landmarks, and, hence, to automatically compute exposure index values.…”

“…4.96 cm in the worst case scenario. Secondly, the EI values associated with the lumbar spine and pelvis are similar and, hence, potential misplacement of c 3 would only slightly affect the exposure indicators (Irrera et al, 2015). Finally, the position of L5 is just roughly inferred from the vertebra T1 and the proximal femur by using a proportionality factor equal to 0.72, i.e.…”

“…It is therefore necessary to detect the anatomical regions of interest that appear in the image. Irrera et al (2015) have recently proposed a landmark-based approach that allows addressing the aforementioned issues. However, the evaluation of the method was conducted from manually annotated landmarks.…”

“…The paper is organized as follows. Section 2 starts by introducing the EI algorithm and by presenting the method for computing EI values with a landmarkbased approach (Irrera et al, 2015). We then present the proposed landmark detection approach and describe our clinical database.…”