Index of content:
Volume 33, Issue 6, June 2006
- Joint Imaging/Therapy Symposium: Valencia B
- Advances in Image‐Guided Interventions
33(2006); http://dx.doi.org/10.1118/1.2241459View Description Hide Description
Algorithmic methods from computer vision and machine learning are dramatically changing the practice of health care and the exploration of fundamental issues in neuroscience. By coupling knowledge of tissue response, atlases of normal anatomy, and statistical models of shape variation, these methods are used to build detailed, patient‐specific reconstructions of neuroanatomical structure from MRI imagery. Such structural models can be automatically augmented with information about function (using fMRI), and about connectivity (using DT‐MRI) to create detailed models of a patient's brain. These models are routinely used for surgical planning — how to reach the target tumor with minimal damage to nearby critical structures; and for surgical navigation — guiding the surgeon to the target site rapidly and safely.
By combining with statistical models of population variation, these methods can also be used to investigate basic neuroscience questions — how different are the shapes of subcortical structures between normal subjects and patients with a specific disease (such as schizophrenia or Alzheimer's); how do these shapes change with development in children, or with administration of pharmaceuticals; how do physiological properties differ between populations (such as the local structure of fiber orientation in white matter tracts). These computational methods provide a toolkit for exploring the structure and connectivity of neuroanatomical structures, in normal subjects and in diseased patients.
This lecture will review current methods for image segmentation and shape analysis.
1. Understand algorithmic methods for image segmentation by tissue type;
2. Understand methods for analyzing shapes of anatomical structures across populations;
3. Understand the application of reconstructedanatomy for surgical guidance and navigation.
33(2006); http://dx.doi.org/10.1118/1.2241460View Description Hide Description
The last decade has seen an explosion in the number of procedures carried out using minimally invasive approaches, as devices and techniques develop. The promise of these new procedures is realized only with a parallel improvement in imaging techniques. Hybrid systems that combine real‐time acquisition with cross‐sectional and/or functional data are one response to the need, and several such systems have now reached the clinic. Depending on the application, different combinations can be used to greater advantage. The interventionalist's familiarity with x‐ray fluoroscopy and the real‐time and projection capability makes it a logical first choice to marry with others such as ultrasound,CT or MR.
The combination of fluoroscopy with C‐arm CT for application to 3D vascular imaging was first implemented in the mid‐90's. New flat‐panel detector technology is now expanding the role of C‐arm CT into low‐contrast imaging applications such as detection of fresh bleeds during treatment of intracranial abnormalities and guidance of vascular and percutaneous abdominal interventions. Highest image quality depends on correction algorithms for lag, truncation, scatter etc. all of which are still under investigation. Recent investigations indicate that contrast resolution as low as 5 HU is possible under clinically realistic conditions.
The combination of fluoroscopy with MR has also been under investigation since the mid‐90's with several prototype long‐travel systems installed in research clinics. While feasibility of such systems has been shown, patient transport between imaging FOVs has been complicated and enthusiasm for the systems has been limited. New flat‐panel detectors which are immune to static magnetic fields, and new x‐ray tube designs are now under development that allow close integration of fluoroscopy and MR systems. Using a double‐donut interventional magnet and a static‐anode x‐ray tube, good image quality can be achieved from both modalities if care is taken to maintain magnetic field homogeneity, rf noise is limited, and the electron optics of the x‐ray tube are controlled. Switching between modalities can take a little as one minute.
The current status of X‐ray/MR, and of X‐ray/CT and C‐arm fluoroscopy/C‐arm CT will be discussed and described. New hardware and software for X‐ray/MR and new image correction and visualization algorithms for fluoroscopy/C‐arm CT will be described. Current challenges in the area will be outlined.
1. Understand the options available for hybrid interventional guidance.
2. Understand the issues related to image quality in C‐arm‐based CT and XMR.
3. Understand the current challenges relating to clinical workflow, multi‐modality registration and image visualization.
The work presented here has been supported by NIH R01 EB003524, P41 RR 09784, R01 EB000198, Siemens Medical Solutions, GE Medical Systems and the Lucas Foundation.
33(2006); http://dx.doi.org/10.1118/1.2241461View Description Hide Description
Magnetic Resonance Imaging(MRI) provides such features as excellent soft tissuecontrast, temperature sensitivity and ability to detect thermally coagulated tissue, that make it well suited for the guidance, control, and monitoring of thermal therapies. In this talk the current progress in MRI guided focused ultrasound thermal ablations and other non thermal interventions will be reviewed. The treatment of uterine fibroids has been approved by the FDA and is on its way to become routine. There are several other clinical trials in testingMRI guided focused ultrasound for the treatment of breast and other tumors. In addition, there is evidence that low thermal exposures controlled by MRI can be used to locally activate treatments. Finally, animal experiments show that ultrasound exposures can induce transient blood‐brain barrier disruption thus allowing targeted delivery of molecules such as chemotherapeutic agents in the brain.
- Functional Imaging for Radiotherapy Guidance
33(2006); http://dx.doi.org/10.1118/1.2241805View Description Hide Description
Brain gliomas are characterized by local infiltration and invasion of surrounding braintissue. The limited responsiveness of these tumors to conventional modes of treatment underscores the critical need to improve our understanding about tumor heterogeneity through the use of in‐vivo imaging, ultimately leading to the design and testing of new treatments. Radiation therapy (RT) is a mainstay in the treatment of malignant braintumors but there is significant room for improvement. Recent technical advances in the field of RT delivery allow for optimized, normal tissue sparing treatments with greater radiobiological effectiveness. However, these new powerful tools can only garner the greatest benefit if directed to the most appropriate (most aggressive and/or radioresistant) tumor region. Magnetic resonance imaging(MRI) is considered the current imaging standard for brain gliomas and is widely used for target definition in RT. However, its information is limited to the morphologic tumor appearance. Radiographically, the presence of contrast enhancement on T1‐weighted images indicates leakage of intravenous contrast into the tumor and signals a disruption of the blood‐brain‐barrier (BBB). This area is currently considered to reflect the most malignant area of the tumor whereas the hyperintensity on T2‐weighted images is presumed to reflect a mixture of edema and tumor cell infiltration. However, it is increasingly accepted that this assumption is not fully justified due to the presence of contrast enhancement in areas of necrosis, the lack of contrast enhancement in certain regions of metabolically active tumor, and the inability of the T2 hyperintensity to distinguish between infiltration and nonspecific processes such as inflammation and reactive edema. Similarly, morphologic imaging is limited in the assessment of treatment effects/response.
New MR‐based techniques have shown promise as a means of providing information on tumor metabolic characteristics and its biological behavior which ultimately will allow us to optimize, monitor, and assess therapeutic interventions beyond that currently provided by tools for the morphologic assessment of a malignant braintumor. 3D Proton Magnetic Resonance Spectroscopy Imaging(MRSI) provides information on tumor cellularity and cell membrane breakdown, cellular energetics, neuronal activity, and hypoxia through its ability to distinguish signals from cellular metabolites such as choline, creatine, NAA, lactate, and lipid. Diffusion Weighted Imaging (DWI) provides additional information on cellularity, cell membrane permeability, intra‐ and extracellular diffusion, and tissue architecture, whereas Perfusion Weighted Imaging (PWI) provides insight into overall cerebral blood volume, tissue microvasculature and vessel permeability. The combination of these metabolic and physiologic modalities with standard anatomic MR modalities will enhance our current understanding of tumor heterogeneity and will provide guidance as to how to optimize current treatment approaches. Based on results from our current studies, we hypothesize that the continued failure of current targeted treatment approaches is in large part caused by insufficient knowledge about the tumor extent, its heterogeneity, and its biological behavior, resulting in directing some or all of the focal therapy to the wrong location.
In addition to assisting in image guidance for RT, these imaging tools hold promise for assessing and predicting therapeutic response and to help distinguish treatment effect and tumor recurrence.
WE‐E‐ValB‐02: Functional Imaging for Radiotherapy Guidance ‐ Quantitative Biological Imaging and the Oncologic Target33(2006); http://dx.doi.org/10.1118/1.2241806View Description Hide Description
Purpose: Biological imaging modalities are reviewed for their quantitative applications, contributions, and limitations in radiation treatment. Concepts are presented for biologically‐matched dose distributions. Method and Materials: Quantitative use and assessment of anatomical and biological oncologyimages shows promise for identification and evaluation of targets for radiation treatment. Biological images, combined with anatomical images, contain digital information on the spatial distribution and intensity level of biological character representative for both cancerous and normal tissues. Quantitative aspects of “bioanatomic” images, for instance, from FDG PET‐CT or MR spectroscopy, enable software processing and manipulation for target localization and delineation, assessment during treatment phase, and post‐treatment evaluation. IMRT provides a method for selective targeting based on biology. Results: Bioanatomic imaging modalities include FDG PET, non‐FDG PET‐CT, MR spectroscopy, diffusion‐weighted MR, MR perfusion, functional MR, magnetoencephalography(MEG), and others. Each modality/technique has finite spatial resolution, contrast (signal‐noise‐ratio), range of voxel intensity values, and sensitivity/specificity. Limitations for quantitative uses include spatial resolution, and “calibration” for validation of voxel intensities and image interpretation. Image digital file formats may be a challenge for certain images. Research opportunities include biology, physics, and imaging science work, and image‐based clinical trials that combine bioanatomic images with advanced dose delivery. Conclusion: Contributions and limitations for quantitative uses of bioanatomical images in radiation treatment are reviewed, including digital characteristics of biological images. Potential benefits include a better understanding of tumor and normal tissue biology and treatment response, and radiation targeting that matches biological conditions.
1. Review characteristics of digital images and the task of image interpretation.
2. Review signal origins relevant to tumor biology for selected imaging modalities.
3. Describe the use of threshold parameters for biological radiation target delineation.
4. Discuss concepts for biologically‐matched radiationdose distributions.
5. List research opportunities for basic science and image‐based clinical trials.
Conflict of Interest: Research sponsored in part by Varian Medical Systems and GE Healthcare.
33(2006); http://dx.doi.org/10.1118/1.2241807View Description Hide Description
Intensity‐modulated radiotherapy(IMRT) is evolving rapidly; the radiation therapy community has begun to regard IMRT as the future standard of care, rather than as experimental or leading‐edge treatment. The effectiveness and benefit of IMRT have been confirmed in multiple cancers, specifically in head and neck and prostate cancers. However, IMRT adoption rates vary widely and are impacted primarily by the learning curve for clinic staff, capacity, and physician's practice patterns. Because physicians must participate to a much higher degree in planning and quality assurance, their learning curves to acquire necessary knowledge and skills in cross section anatomical images are also steep. Image‐guidedradiotherapy(IGRT) or adaptive radiotherapy is an emerging radiation therapytreatment methodology that complements IMRT.IGRT takes tumor and organ motion into consideration and monitor this motion in a near real time mode to make sure radiation is being delivered to the tumor. Furthermore incorporating functional images into management decision‐making has highlighted the future perspective of IGRT. However, to radiation oncology community the explosion of overwhelming technology has created challenges which include the lack of enabling tools to facilitate bench to bedside research, bridge knowledge gap, and improve quality and operational efficiency.
33(2006); http://dx.doi.org/10.1118/1.2241808View Description Hide Description
Radiation therapy is an image‐guided procedure whose success depends strongly on the image modality used for treatment planning and the level of integration of the available imaging information. Advancement in IMRT has provided an unprecedented means to produce highly conformable dose distribution while sparing sensitive structures, which calls for better imaging tools for tumor target definition and for the management of inter‐ and intra‐fractional organ motion. In this talk we will summarize recent advances in functional and molecular imaging techniques and discuss various issues related to the integration of the newly emerged imaging data into radiation therapy planning. It is anticipated that the new imaging modalities will play an important role in radiation oncology practice and make significant impact in cancer diagnosis, staging, treatment planning, and monitoring of therapeutic response. The potential impact of biologically conformal radiation therapy (BCRT) or biologically guided radiation therapy (BGRT) will be discussed. Finally, issues related to the quality assurance of functional and molecular imaging and BCRT will also be addressed.
1. Introduce the concept of functional and molecular imaging.
2. Illustrate the steps involved in integrating molecular imaging such as PET and MRSI into treatment planning process.
3. Introduce PET/MRI/MRSI and CTimage fusion techniques (including deformable image registration).
4. Provide an overview on recent advances in PET/CT and MRSI, and update on the development of new PET tracers and data acquisition techniques.
This work was supported in part by NCI 5R01 CA98523‐01.
- Imaging the Tumor Micro‐environment and Response to Therapy
33(2006); http://dx.doi.org/10.1118/1.2241701View Description Hide Description
The reasons for the failures of conventional anticancer therapies are varied and multiple. It is however, becoming increasingly clear that physiological conditions in tumors, arising primarily from inadequate and non‐uniform vascular networks, can play significant roles in the lack of therapeutic responsiveness of neoplasms. However, the impact of the tumor microenvironment far exceeds its direct effects on therapeutic treatment modalities. For example, tumor hypoxia can contribute to processes that directly favor malignant progression and increased metastasis. New treatment strategies aimed at improving tumor response through targeting cells existing in these microenvironments or the aberrant tumor vasculature itself, therefore are of high interest. Non‐invasive imaging strategies may aid not only in identifying tumors that would be effectively treated with such strategies but also in providing effective means of monitoring treatment effectiveness and therapeutic outcomes.
WE‐C‐ValB‐03: Functional Analysis of Tumor Vasculature Post Antiangiogenic Intervention with Argeted Anti‐VEGFR2 Therapy33(2006); http://dx.doi.org/10.1118/1.2241702View Description Hide Description
High frequency ultrasound biomicroscopy (UBM) permits in vivo assessment of anatomical and physiological parameters at high resolution. The ability to image the mouse noninvasively facilitates longitudinal studies with repeated measurements in the same animal over time, to follow normal development or disease processes as well as response to therapeutic intervention. Anatomical images are obtained using B mode, giving a cross‐section of the structure of interest. Anatomical changes can now be quantitated using 3D volumetric B mode imaging, and rendered to give an accurate contour of the structure of interest. 3D UBM may be used to measure organ growth through normal development, or to compare treated versus untreated populations in mouse models of human disease. High frequency ultrasound Doppler permits measurement of blood flowvelocity in selected vessels of interest. Recent advances in blood flow analysis, including speckle variance analysis, permit investigation of blood flowvelocity of flowvelocities down to 100 microns per second, as well as perfusion alterations.
High frequency ultrasound imaging modes, imaging strategies, and post‐processing methods used to generate anatomical and physiological data will be discussed.
1. Understand the concepts behind high frequency ultrasound imaging modes.
2. Understand the application of high frequency ultrasound to image development and disease processes in the mouse, including experimental design, data acquisition and data processing.
3. Understand therapeutic study design, data acquisition and analysis using UBM.
WE‐C‐ValB‐04: Magnetic Resonance Spectroscopy Detects Metabolic Changes Upon Chemotherapy of Human Non‐Hodgkin's Lymphoma Xenografts33(2006); http://dx.doi.org/10.1118/1.2241703View Description Hide Description
A preliminary multi‐institutional study has recently demonstrated that ratios of the hosphomonoesters, phosphoethanolamine plus phosphocholine, to NTP of human non‐Hodgkin's lymphoma (NHL) easured by 31P MRS before initiation of therapy can identify about 2/3 of the patients who will not exhibit a complete local clinical response. These patients should be encouraged to undergo more aggressive alternative therapy, which entails some risk of mortality but also offers hope of response, remission or cure. However, the limited sensitivity of 31P MRS limits its application mostly to large, superficial tumors. We have, therefore, been developing much more sensitive 1H MRS and MRI methods to examine smaller tumors. Here we report studies of mouse xenografts of the most common form of human NHL and the only form that exhibits cures (in about 1/3 of the patients) — diffuse large B‐cell lymphoma (DLCL2). In vivo 1H MRS using a selective multiple quantum coherence pulse sequence (Sel‐MQC) detected decreases of lactate and total choline in DLCL2 tumors treated with three cycles of CHOP (cyclophospamide, doxorubicin, vincristine, prednisone) chemotherapy with Bryostatin added to suppress expression of the mdr1 gene. Single voxel localized spectroscopy (STEAM) detected decreases in choline and lactate/lipid that accompanied response. These changes correlated decreases in phosphomonesters detected by 31P MRS with tumor growth delay. In vivo data were correlated with data on tumor extracts. Our data suggest that 1H MRS may provide a very sensitive method for detecting early response of human NHL to chemotherapy; 1H MRI studies of response are in progress.
- Tomographic Guidance of Radiotherapy Procedures
33(2006); http://dx.doi.org/10.1118/1.2241555View Description Hide Description
Mega‐Voltage Cone‐Beam Computed Tomography (MVCBCT) systems have been in clinical use at UCSF since February 2005. This lecture will provide a general description of the MVCBCT system, image acquisition,reconstruction and registration. The practical clinical applications as well as the guidance strategies will be presented through a number case studies including the following anatomical sites; head&neck, lung, spine, prostate and pelvic sites.
1. Understand the basics concepts of MV Cone‐Beam CTimaging.
2. Understand the range of clinical applications made possible by MVCBCT in the following categories:
a. Patient and organ positioning
b. Monitoring of anatomical changes and tumordose response
d. Dosimetric impact
e. Dose planning in presence of CT non‐compatible objects
g. Dose‐guided radiation therapy (DGRT).
This work was supported by Siemens Oncology Care Systems.
33(2006); http://dx.doi.org/10.1118/1.2241556View Description Hide Description
Purpose: To account for geometrical uncertainties during radiotherapy,safety margins are applied. In many cases, these margins overlap organs at risk thereby limiting dose escalation. The aim of image‐guidedradiotherapy is to improve the accuracy by imagingtumor and critical structures on the machine just prior to irradiation. NKI collaborated in the development of a kV cone beam CT guided accelerator. Method and Materials: The chosen imagingdose is 3 cGy for prostate, 1 cGy for head and 2 cGy for lung, 4D scanning. The availability of high quality tomographic images and automatic image registration on the machine leads to many new clinical applications, such as high precision hypofractionated treatments of brain metastases and solitary long tumors with on‐line tumor position corrections. Adaptive radiotherapy(ART) of prostate cancer is now also in routine use. We adapt to the average prostate and rectum using cone beam scans made during the first week of treatment. Results: The prostate is located automatically in 85% of prostate scans. Even though we use laxatives, the main confounding factor is short‐term mobility due to moving gas that causes streak artifacts in the CT reconstructions. Our ART protocol allows reducing the margin from 10 to 7 mm. Patient localization with 1 mm accuracy (bony anatomy) is easily achieved with the current equipment. Pre‐ and post‐treatment scans demonstrate negligible motion of about 0.5 mm SD, both for brain and bladder cancer patients. Conclusion: The availability of cone beam CT on the linear accelerator makes ART very efficient and more accurate, since problem duplicating the setup on the CT scanner are avoided. For all image‐guided protocols, the residual uncertainties need to be taken into account, and the safe level of margin reduction evaluated. In conclusion, kV cone beam CT guided radiotherapy is now very much a clinical reality.
1. Understand strengths and weaknesses of kV cone beam CT guided radiotherapy.
2. Understand which clinical protocols are most suitable for this technique.
3. Understand the necessity of careful uncertainty analysis and margin selection, especially with image guidance.
This research was partially sponsored by Elekta Oncology Systems.
33(2006); http://dx.doi.org/10.1118/1.2241557View Description Hide Description
Magnetic Resonance Imaging(MRI) has become the gold standard in the imaging of many soft tissues, osseous and joint conditions, but to date there has been limited success in harnessing this excellent imaging modality for interventional procedures. MRI is an ideal interventional guidance modality: it provides real‐time high‐resolution images at arbitrary direction and is able to monitor therapeutic agents, surgical tools, biomechanical tissue properties, and physiological function. At the same time, MRI poses formidable engineering challenges by limited access to the patient and a strong magnetic field that prevents the use of conventional materials and electronic equipment. Currently, no exigent technical solution exists to assist MRI‐guided needle placement procedures in an accurate, simple, and economical manner.
A wide variety of procedures may be performed on open magnets but the trend is to use high field closed magnets, mainly because of improved imaging quality and wider availability of pulse sequences. The higher the field the higher the SNR and the higher SNR can be used to improve spatial and temporal resolution and can make techniques like temperature or flow sensitive imaging, functional brain MRI, diffusion imaging or MR spectroscopy more useful. Considering these trends, we believe that the use of conventional high‐field closed MRI scanners for guidance will allow more successful dissemination of MR‐guided techniques to radiology facilities throughout the country and eventually beyond.
The talk will survey major trends and achievements in MR compatible interventional robotics, and present specific research projects and results currently in progress at the Johns Hopkins University and collaborators.