Digital tomosynthesis system has been widely used in chest, dental, and breast imaging. Since the digital tomosynthesis
system provides volumetric images from multiple projection data, structural noise inherent in X-ray radiograph can be
reduced, and thus signal detection performance is improved. Currently, tomosynthesis system uses two data acquisition
modes: step-and-shoot mode and continuous mode. Several studies have been conducted to compare the system
performance of two acquisition modes with respect to spatial resolution and contrast. In this work, we focus on signal
detectability in step-and-shoot mode and continuous mode. For evaluation, uniform background is considered, and eight
spherical objects with diameters of 0.5, 0.8, 1, 2, 3, 5, 8, 10 mm are used as signals. Projection data with and without
spherical objects are acquired in step-and-shoot mode and continuous mode, respectively, and quantum noise are added.
Then, noisy projection data are reconstructed by FDK algorithm. To compare the detection performance of two
acquisition modes, we calculate task signal-to-noise ratio (SNR) of channelized Hotelling observer with Laguerre-Gauss
channels for each spherical object. While the task-SNR values of two acquisition modes are similar for spherical objects
larger than 1 mm diameter, step-and-shoot mode yields higher detectability for small signal sizes. The main reason of
this behavior is that small signal is more affected by X-ray tube motion blur than large signal. Our results indicate that it
is beneficial to use step-and-shoot data acquisition mode to improve the detectability of small signals (i.e., less than 1
mm diameter) in digital tomosynthesis systems.