The future of ultrasound: viewing the dark side of the moon?

As our field has matured, other cross-sectional imaging methods have emerged such as magnetic resonance imaging (MRI). MRI has made inroads into the applications of ultrasound and, in some cases, has even challenged ultrasound's position as the first and most relied upon approach for diagnostic imaging. MRI and computed tomography (CT) have the appeal of being more standardized than ultrasound and thus less operator-dependent. Since two-dimensional (2D) ultrasound is akin to fluoroscopic techniques, ultrasound has always suffered from the need for specialized training and skills—not only for interpreting the scans but also in producing optimal images. Thus, current 2D ultrasound has been hard to standardize, due to its limitations based on operator experience and single-slice imaging dictated solely by manual transducer positioning. MRI is far more consistent, typically producing images that are read by someone other than the person who generated them, thus placing far more reliance on the actual technology than on operator technique. Since MRI images are manipulated in reconstructed planes after completion of the examination, MRI also requires far less physician time with the patient. The advent of three-dimensional (3D) volume imaging is one of the most important advances in modern sonography1, 2. While MRI is certainly proving valuable, the maturation of 3D ultrasound will go a long way towards matching MRI's capabilities. 3D ultrasound volume imaging has enabled the ultrasound practitioner to obtain volumes that can be reconstructed in any plane after the patient has left the clinic; this capability will likely promote the development of standardized ultrasound techniques, akin to what MRI has offered. Thus, 3D ultrasound and the reconstruction of multiplanar slices will probably permit ultrasound to solidify its position in relation to MRI in body imaging. 3D volume imaging promises to offer the standardization that ultrasound needs in order to compete with other disciplines such as MRI and CT in simplicity, reliability and standard approach to diagnosis. Ultrasound will be able to move from the artistic to the standard planes of imaging, which are understandable by many more practitioners. Let us consider the example of real-time scanning of the fetal heart in three orthogonal planes simultaneously. This new techniques facilitates the evaluation of the fetal heart, both normal and abnormal, by providing standard views regardless of fetal position3. This type of virtual cardiac examination is far more informative than a videotape, which is entirely dependent on the skill of the operator. Furthermore, the volumes obtained can be sent remotely for evaluation by an expert, even if the operator did not fully understand the cardiac anatomy4. What is 3D volume imaging likely to do for ultrasound of the female reproductive tract in the coming decade? Many reports have established the advantages of 3D ultrasound in fetal imaging. I will confine my comments here to the exciting and previously relatively unexploited field of general sonography in the female pelvis. The pelvic images obtainable using 2D ultrasound are limited to those available from particular portals—either the anterior abdominal surface or the vagina. These windows restrict the orientation from which images can be generated and increase the operator skill, some might say artistic ability, needed to generate readable pictures. It is quite possible that standard 2D images generated from the point of view of the vagina or anterior abdominal surface may not produce the optimal plane for imaging of the pelvis or fetus. We have never had the opportunity to aim the ultrasound beam from the hip or through the buttocks of the patient to see if the anatomy is better displayed. We have, therefore, lacked a basis of comparison to determine whether the customary image planes that we have always used by necessity are, in fact, the best ways to image these organs. Are the most favorable planes to image the female pelvis the ones directly available from the vaginal orientation or have we made do with the only options technology has offered thus far? Now, 3D reconstruction enables us to obtain an image slice from any orientation, including the millions of planes inaccessible through the standard 2D portals. This advance permits us, for the first time, to ‘view the dark side of the moon’, albeit indirectly. I am convinced that this new flexibility will provide many more heretofore hidden imaging capabilities in ultrasound. Eventually we may find that the entire pelvis will be better scanned in a plane other than the one traditionally used, a notion that we could not consider prior to the advent of 3D ultrasound. As technology improves and as our colleagues in the ultrasound industry develop new instruments and software, we will be able to scan in real time in any reconstructed plane. We will be able to explore the sonographic anatomy in many planes not yet investigated. Although such reconstruction generally entails a loss of resolution compared to the original planes, I am certain that technology will improve rapidly such that any plane will soon be available for real-time ultrasound imaging, and with acceptable resolution. We will then be able both to choose the most appropriate scanning plane and to overcome the very narrow orientations available through the traditional portals. This ability to image in any plane opens up an enormous capability that ultrasound has previously lacked and that other types of cross-sectional imaging have enjoyed. 3D reconstruction will also give ultrasound the ability not only to image in real time but also to become just as standardized as MRI with its uniform interval planes. Thus ultrasound may become less dependent upon the skill of the operator. New transducers, which standardize the sweep to produce a 3D volume set, will become the norm, since hand-held, 2D transducers have no fixed frame of origin and require the skill of the operator for spatial location and orientation. The skill needed to perform 2D ultrasound today requires a great deal of training to obtain 3D information from a 2D display. Reconstruction of 3D volumes will make it possible actually to display this information, thus demanding less specialized training and less physician time. Physician time can then be applied to manipulation of those views into a far greater array of possibilities than ever before. Ultrasound will retain the ability to image in real time, which gives sonography a huge degree of flexibility in its applications not yet available with MRI. Furthermore, in the process of standardizing ultrasound using volume reconstruction, a feasibility study has shown that a virtual ultrasound examination can be done at one site and the 3D packets of information can be sent to another site where the diagnosis can be made5. Such virtual patient examinations done at one site and interpreted at another will become realistic. 3D is not new: CT and MRI have employed this phenomenon for several decades with reconstruction of volumes in many orthogonal planes and displaying the images in soft tissue or bony windows. Now that ultrasound has acquired similar capabilities, the benefits of using ultrasound as the first imaging test are emerging rapidly and are allowing ultrasonography to solve new problems. For example, real-time imaging of the coronal view of the uterus to determine uterine shape is quick and simple, thus providing valuable information to patients with recurrent miscarriage and infertility (Figure 1). Until recently, MRI was the only cross-sectional imaging modality that could display the shape of the uterine cavity and diagnose a septate versus a bicornuate uterus6. Now, reconstructed planes of 3D ultrasound images have supplanted the use of MRI, not only because of ultrasound's similar capabilities in imaging but because it is readily available, cheaper and faster7. Thin slice (2 mm) C-plane view (VCI) of the uterus on split screen in real time. Note the septate uterus. The entire uterus can be interrogated in this plane. Imaging the floor of the pelvis has also been an area dominated by MRI, and even endovaginal MRI8-10. Ultrasound now permits similar images with just a quick reconstruction of a 3D translabial volume to demonstrate the levator ani, anus, rectal sphincter, position of the urethra and vagina (Figure 2)11. Perhaps as we investigate this new role of ultrasound, 3D ultrasound will become the first approach used in the evaluation of patients with urinary incontinence, pelvic floor or anal symptoms. Multiplanar scan of the pelvic floor imaged from the perineum (a). Note that the C-plane depicts the coronal view of the pelvic floor (b) akin to a magnetic resonance imaging slice. Imaging of the adnexa in the standard transvaginal 2D scan often presents us with the problem of trying to decide whether a septate adnexal cyst represents a hydrosalpinx or a complex ovarian cyst. Using real-time ultrasound, one can activate planes outside the scanning plane and actually orient the septate cyst in such a way as to see whether or not these cysts connect. This maneuver enables us quickly and easily to diagnose or exclude a hydrosalpinx (Figure 3). Small postmenopausal ovaries sonographically often elude standard 2D techniques. These ovaries can be small, comma-shaped objects that may be better visualized in a non-acquisition plane. In these cases I have found a real-time split screen helpful, with a standard plane on one side and the C-plane on the other, often producing a reconstructed image that renders the ovary more obvious (Figure 4). Thin slice (2 mm) C-plane view (VCI) of the adnexa on split screen in real time. The standard view shows a septate cyst. The VCI plane can be manipulated in real time to demonstrate the hydrosalpinx. Hard-to-see postmenopausal ovary is far better visualized on the C-plane (VCI), which was manipulated in real time to find the ovary. It is now up to the ultrasound community to discover all of the areas that have become accessible to ultrasound due to this ability to reconstruct any plane and scan it in real time. This progress allows us to maintain ultrasound's role in cross-sectional imaging applications. I am confident that eventually both modalities, i.e. 3D ultrasound and MRI, will contribute to the field of obstetrics/gynecology imaging and patient care. By virtue of the advances described here, however, ultrasound should remain the first line of the diagnostic armamentarium in cross-sectional imaging. Another area of incredible advancement has been the miniaturization and portability of a whole generation of ultrasound instruments in development. These instruments may become so small as to be likened to a hand-held personal digital assistant, and be carried in the pocket much like a stethoscope. Several editorials in the ultrasound and cardiology literature have used the term ‘sonoscope’, and suggested that a limited examination may become a routine extension of the physical examination being used by doctors as well as medical students12, 13. In other countries—Germany in particular—medical students are already being trained in the use of bedside ultrasound as part of a patient assessment13. This would certainly not eliminate the need for a full ultrasound examination, which will still require the same expertise it does today (and may include 3D ultrasound). Ultrasound systems, however, can become so miniaturized and so crucial to the physical diagnosis that they will inevitably become more widespread in the medical profession than they are today. Currently, the limited abdominal scan is utilized by many emergency departments as the first bedside evaluation after abdominal trauma in the search for free peritoneal fluid as a sign of organ damage14. As the technologies become smaller and more portable it is likely that this type of application will expand. This change in practice will obviously result in many more referrals to the fully equipped ultrasound departments for further expert evaluation. As technology advances in so many ways, the physical examination may start to incorporate ultrasound as an extension of a physician's inspections, palpation and auscultation. Currently there are guidelines and standards in the United States that govern the practice of diagnostic ultrasound. Clearly the miniaturization of ultrasound devices will require a review of these standards and procedures for the limited screening scans. Ten years from now, I can imagine speaking with young ultrasound specialists and reminiscing about the old days when ultrasound displays were limited only to the acquisition planes. I will also tell them about the era when ‘portable’ ultrasound systems had to be wheeled from room to room, rather than carried in a pocket! With the development of 3D volume scanning, as well as the miniaturization and portability of tiny ultrasound scanners, ultrasound has the opportunity to make a quantum leap forward, thus cementing its position in standard cross-sectional imaging in addition to extending its availability to far more providers in the screening arena. These advances promise to ensure the longevity of medical ultrasound and anchor its role in the world of diagnosis.