Comparison of Dual-Energy Subtraction and Electronic Bone Suppression Combined With Computer-Aided Detection on Chest Radiographs: Effect on Human Observers' Performance in Nodule Detection

Abstract

very few series have dealt with a combination of these techniques [14,15]. The feasibility of software-only bone suppression on conventional chest radiography in combination with CAD has not been tested to date. Furthermore, of the studies in the literature, most investigators selected image collections in which each image contained only one lesion, a scenario that often does not correspond to the patient collective encountered during daily routine.The aim of the current investigation was to compare the effect of dual-energy subtraction and recently developed bone suppression software alone and in combination with CAD on the diagnostic performance of human observers in lung nodule detection. Materials and MethodsThis retrospective study was performed according to the regulations of the institutional review board and required no informed consent. C onventional chest radiography is still the most frequently performed initial imaging evaluation used worldwide to exclude pulmonary nodules because of its wide accessibility, low cost, and low radiation dose. In past decades, several technical innovations were introduced to improve the sensitivity of this modality. In addition to dual-energy subtraction radiography, other new software techniques have been shown to be capable of decreasing or completely removing the overprojection of bones, mainly of the clavicles and ribs, that significantly hampers the detection of pulmonary nodules on projection radiographs [1][2][3][4][5][6]. On the other hand, computer-aided detection (CAD) techniques have been developed for chest radiography, analogous to those used with chest CT, to aid in the detection of lung nodules. Studies have analyzed the diagnostic gain from bone subtraction or suppression and CAD separately [7][8][9][10][11][12][13] Z. Szucs-Farkas received fees from the Riverain Technologies Group for collecting anonymized patient data for a different reader study. That study is not related to the current investigation in any way. Comparison of Dual-Energy C a r d io p u lm o n a r y I m ag i ng • O r ig i n a l R e s e a rc hOBJECTIVE. The objective of our study was to compare the effect of dual-energy subtraction and bone suppression software alone and in combination with computer-aided detection (CAD) on the performance of human observers in lung nodule detection. MATERIALS AND METHODS.One hundred one patients with from one to five lung nodules measuring 5-29 mm and 42 subjects with no nodules were retrospectively selected and randomized. Three independent radiologists marked suspicious-appearing lesions on the original chest radiographs, dual-energy subtraction images, and bone-suppressed images before and after postprocessing with CAD. Marks of the observers and CAD marks were compared with CT as the reference standard. Data were analyzed using nonparametric tests and the jackknife alternative free-response receiver operating characteristic (JAFROC) method.RESULTS. Using dual-energy subtraction alone (p = 0.0198) or CAD alone (p = 0.0095) improved the detection ra...

Computer-Aided Nodule Detection System

“…A recent study reported that use of OnGuard, version 5.1, can significantly improve the sensitivity of human observers for the detection of 5-to 29-mm lung nodules in a selected patient data set (12). A more recent study reported that the same CAD system used in the present study improved radiologists' performance in the detection of solitary lung nodules on 111 PA CXRs, even when baseline performance was optimized by providing lateral radiographs and BSI images (13).…”

“…Software BSI techniques (9)(10)(11)(12)(13)(18)(19)(20) can retrospectively create an image from a standard CXR that is similar to the soft-tissue image generated by a dedicated DES radiography unit. In a previous study with an early version of the CAD system, the CAD detection rate was 46% with 2.4 FPs per image on standard CXRs compared to 67% with 1.6 FPs per image on DES soft-tissue images (21) generated by a dedicated DES unit; however, such DES devices are not routinely available in clinical practice.…”

“…Until now, the performance of CAD systems applied to CXRs has been evaluated almost exclusively with small numbers of highly selected cases, which typically included a limited number of nodules in otherwise abnormality-free lungs (6)(7)(8)(9)(10)(11)(12)(13). To simulate the effect of using CAD in routine clinical practice, this study investigated CXRs from patients with and without lung diseases in an otherwise unselected series.…”

Improved Lung Cancer Detection in Cardiovascular Outpatients by the Pulmonologist-based Interpretation of Chest Radiographs

“…There are many methods to improve detection of lung nodules on chest radiographs such as computer-aided diagnosis, and the removal or suppression of bony and other anatomical structures (28)(29)(30)(31)(32)(33). However, the clinical usefulness and cost-benefit of these techniques for detecting lung cancer remains controversial.…”

Radiology as an aid to diagnosis in lung disease

“…As a result, there are far fewer poor-quality CXRs produced nowadays. Capturing CXRs in digital form has enabled further advances such as three-dimensional (3D) imaging using two or more spatially separated views, tomographic imaging using multiple views to generate one data set from which slices through the chest can be imaged individually and dual energy images to enable the separation of bony structures from soft tissues 2 3. These advances are yet to become standard practice, but are commercially available, and the masking of bony structures, particularly the ribs, has facilitated the development of computer-aided diagnosis for lung nodules.…”