Quantitative Medical Imaging

The field of medical imaging is undergoing a radical shift, from subjective interpretation to quantitative analysis and measurement. This transformation is well established in the clinical trials arena, and is beginning to enter the diagnostic field as well. This paper will consider the implications of this change in terms of instrumentation, procedure, and analysis. It will include a brief review of the principal imaging modalities (CT, MRI, PET, Ultrasound), and will examine in detail the acquisition and analysis techniques that will be required to successfully complete the transition from qualitative to quantitative imaging. Introduction In the past three decades, medical science has made great strides toward the understanding of cellular biochemistry and the mechanisms of disease. In concert with this change, the technologies available for medical imaging have proliferated. In particular, cross-sectional imaging techniques, which allow the precise reconstruction of three dimensional structures within the body, and functional imaging techniques, which allow the assessment of biological function as well as form, have become widely available. These imaging techniques potentially provide a tremendous amount of information about disease state and response to treatment. However, the means to interpret that information has badly lagged behind the ability to acquire it. A good example of this phenomenon is given by the RECIST (Response Evaluation Criteria In Solid Tumors) standard, which is the primary imaging endpoint for assessing disease progression or response to treatment in many types of cancer. This technique reduces the assessment of structural changes in tumors to a simple summation of longest diameters, limited to the axial imaging plane. This technique was originally developed with plane film x-ray imaging in mind, and fails to take advantage of the far richer three dimensional information set available in a spiral CT scan. In some cases, the assessment gleaned from RECIST will correlate well with the changes seen in “true” tumor volume. In others, however, it will not. An example of this is given in Figure 1 below. The continued use of RECIST as a standard evaluation in cancer is not due to any strong argument that single diameters are better than, or even equivalent to, a full volumetric assessment. Rather, it is due to the lack of availability of convenient and reliable tools for producing volumetric measurements. Several groups, including VirtualScopics, are now in the process of producing, validating, and commercializing tools in this and other areas which will allow the use of precise, quantitative measurement in medical imaging, first in the clinical trials arena, and later in the clinic itself. Figure 1: (l) Baseline scan showing a large tumor filling much of the subject’s right sinus cavity. Standard RECIST measurement is shown by black arrows. (r) Follow-up scan showing the same tumor, which has cavitated, losing approximately 70% of its bulk. However, the standard RECIST measure is roughly unchanged. Quantitative Medical Imaging in Clinical Trials Recent scientific advances have brought about massive changes in the business of drug discovery and development. Large companies that once had relatively few compounds in development now juggle hundreds of potential candidate compounds. For the industry to sustain itself, companies must devise ways to quickly identify promising compounds and cull ineffective ones. Various estimates place the cost of developing a single compound from discovery to the pharmacy shelf at between $800 million and $1.7 billion. The Tufts Center for the Study of Drug Development in its Impact Report, Volume 4, Number 5, September/October 2002 estimated that it costs $808 million, on average, to develop and win market approval for a new drug in the United States. This study stated that better pre-clinical screens, to increase success rates from the current 21.5% to one in three, could reduce capitalized total cost per approved drug by $242 million. Quantitative imaging can reduce late-stage attrition dramatically by offering more accurate information much earlier in the drug development process. As an example, dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) is able to provide information about blood flow and vascular permeability in tumors. This information allows a relatively small trial to quickly determine whether an anti-angiogenic or vascular disruptive agent is having the desired effect [1,2]. Whereas Phase I clinical trials normally focus on dosage and safety, studies that incorporate quantitative imaging can also test for drug efficacy, offering information that can save millions of dollars by allowing companies to better prioritize their drug pipelines and make go/nogo decisions much earlier. In pre-clinical research, scientists can test method of action and lay the foundation for a more streamlined clinical trial process from Phase I through Phase III by obtaining critical information about efficacy. The value of medical image analysis stems from quantification and automation. While the primary shortcoming of standard endpoints, such as pain or functionality scoring, is that they are largely subjective and difficult to reproduce, quantitative imaging allows the replacement of a subjective evaluation — knee pain ranked on a scale of 1 to 10 — with an objective quantification — cartilage volume in cubic millimeters. Automation in the image analysis process — using a computer algorithm to measure lesion size rather than a clinician with a ruler — provides a degree of accuracy and reproducibility that cannot be duplicated by manual techniques. A good example of this phenomenon is provided by the measurement, using MRI, of lesion burden in multiple sclerosis (MS) patients. MS lesions generally are irregularly shaped, and tend to have fuzzy, indistinct boundaries (see Figure 2). As a result, several studies have estimated the inter-operator coefficient of variability (CV) in white matter lesion burden measurement at 20% or more [3,4] and the intra-operator CV at ~7%. Introducing automation into this process can reduce this variation to 2% or less [3], allowing statistically significant efficacy findings to be obtained far earlier in the development process. Figure 2: A T2 weighted MRI scan of the brain of a multiple sclerosis patient. The irregular bright areas surrounding the ventricles are white matter lesions. The small size and indistinct boundaries of these lesions make them very difficult to quantify manually. Precise, automated measurement brings another critical benefit: it enables the detection of small changes in structure and function over time. In evaluation of osteoarthritis, for example, MRI of the cartilage in the knee coupled with automated measurement of volume and chemical composition shows disease changes in months; these changes would not be apparent using standard x-ray evaluation for years. With this quality of information, researchers can more confidently make the go/no-go decision for a compound early in the evaluation process, allowing scarce resources to be allocated to the most promising candidates. Reproducible medical image analysis is driven by algorithms that enable quantitative, volumetric measurement of structures and metabolic functions. Guided by the information present in the images, as well as embedded anatomical knowledge, the algorithms enable segmentation of different tissue types such as bone, muscle, fat and fluid. From an MRI knee scan, for instance, it is possible to produce a three-dimensional reconstruction that graphically distinguishes cartilage from underlying bone, as well as from ligaments, fluid, degenerated menisci or inflamed synovium (see Figure 3). This capability provides a valuable assessment tool for clinical research in osteoarthritis — a disease with multiple endpoints — because it allows the very sensitive and specific measurement of all the components of the knee joint and the detection of small changes in any of those components over time. Figure 3: Three dimensional rendering of the knee joint of an osteoarthritis subject, showing tibial, femoral and patellar cartilage