Interpreting medical ultrasound images is a challenging task, requiring technicians to mentally reconstruct 3D representations from 2D images. To simplify this, MIT researchers developed a new ultrasound imaging approach that allows users to visualize a 3D augmented-reality image of the scanned object. This technique could expedite the training process for ultrasound technicians and enhance accuracy in clinical settings, such as accurately placing needles for biopsies. Canan Dagdeviren, an associate professor at MIT, states, "This could make ultrasound more intuitive and understandable. It could also be less time-consuming and provide healthcare providers with peace of mind."
Ultrasound imaging works by bouncing high-frequency sound waves off tissues, which are then converted into 2D images by an ultrasound transducer. Technicians are trained to convert these images into 3D mental representations, a skill that is difficult to master and has a long learning curve. The MIT team proposed to reduce cognitive load by combining 3D ultrasound imaging with augmented reality (AR).
For this study, a real-time 3D system was developed for breast cancer detection, featuring an ultrasound probe slightly smaller than a deck of cards, utilizing a chirped data acquisition system (cDAQ). The probe's ultrasound array is configured in a hollow square shape, enabling it to capture 3D images of the underlying tissue. This system, requiring fewer ultrasound elements, is cheaper and consumes less power. The data collected can be compressed and streamed into a 3D graphics engine, Unreal Engine, which converts voxel data into a direct 3D representation without loss of information.
Wearing an AR/VR headset, users can see a 3D rendering of the internal structure superimposed over the actual object, akin to having X-ray vision. By tilting their heads or approaching from different angles, users can view the object from various perspectives, facilitating identification.
The researchers tested their AR-VIU (augmented real-time volumetric imaging in ultrasound) technology with 18 participants, finding that it significantly improved users' ability to identify and locate objects, especially for novices who performed nearly as well as experts. Most novices preferred the AR-VIU approach, stating it made tasks easier.
The researchers are now working to improve imaging resolution and conduct further tests to validate the accuracy of the AR-VIU technology. This research was funded by the MIT Media Lab Consortium, the National Science Foundation, an MIT HEALS graduate fellowship, and an MIT-Tata graduate fellowship.
Blogger's Review: By integrating augmented reality with ultrasound imaging, MIT's research not only enhances the efficiency of medical image interpretation but also significantly eases the learning curve for novices. This innovative application deserves broader promotion in medical settings and could potentially transform training and clinical practices in ultrasound technology.