Sensory Perception & Interaction Research Group

With the advent of digital imaging, the number of images that radiologists must view to make a diagnosis has grown from around 100 to thousands [4]. Meanwhile, the user-input options available for diagnostic purposes have remained relatively unchanged and are likely limiting workflow efficiency [4]. Radiologists also experience a high rate of repetitive stress injuries from prolonged use of the workstation [1]. Innovations allowing images to be read more quickly without reducing accuracy or thoroughness, and improving ergonomics at the same time, would increase efficiency and save money. Errors in diagnosis are of concern to any medical system, and can be caused by fatigue and errors in perception. Fatigue occurs over the workday and was seen to cause a small drop in detection of fractures in radiographic images between early and late in the day [3]. Subjective radiologist feedback related this performance drop to increased physical discomfort, sleepiness, lack of energy, and oculomotor strain; the eyestrain effect is worse for CT images [8]. Three causes of false-negative type perception errors have been observed [4]: never fully fixating on the object, not fixating long enough to recognize the object, and fixating but not recognizing or consciously dismissing the object. Radiologists primarily use mouse and keyboard to interact with images. But are they used purely because of familiarity and lack of alternatives? The mouse was not designed for continuous use and therefore can cause repetitive stress injuries [2], as the mouse’s restricted action set requires PACS interactions to be modal and repetitive. Many consumer devices have therefore been investigated as alternative inputs [9]. For instance, a pen and tablet reduces the contouring time for a ‘region of interest’ compared to a mouse and keyboard, likely due to a reduction in contouring errors [6]. Overall, however, the mouse generally outperforms other devices in pointing performance [2]. If the mouse interaction was more intuitive, could it decrease the mental strain that is likely contributing to the aforementioned fatigue errors? In parallel, could enhanced mouse interaction improve visual search performance, thereby reducing perception errors?