BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online July 19, 2006
British Journal of Radiology (2007) 80, 113-120
© 2007 British Institute of Radiology
doi: 10.1259/bjr/36793733
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by White, M J
Right arrow Articles by Totman, J J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by White, M J
Right arrow Articles by Totman, J J

Full paper

Parametric mapping of the hepatic perfusion index with gadolinium-enhanced volumetric MRI

M J White, MPhys1,6, R L O'Gorman, MSc2,3, E M Charles-Edwards, MSc1, P A Kane, MBBS, MRCP, FRCR4, J B Karani, BSc, MBBS, FRCR4, M O Leach, PhD, CPhys, FMedSci1 and J J Totman, MSc5

1 Cancer Research UK Clinical Magnetic Resonance Research Group, Royal Marsden NHS Trust & Institute of Cancer Research, Downs Road, Sutton, Surrey SM2 5PT, Departments of 2 Neuroimaging, 3 Medical Engineering and Physics and 4 Radiology, King's College Hospital, London SE5 9RS, 5 Brain & Body Centre, University of Nottingham, Nottingham NG7 2RD, 6 Centre for Medical Image Computing, Department of Medical Physics, UCL, London WC1E 6BT, UK

Correspondence: Mr Mark J White, Centre for Medical Image Computing, Department of Medical Physics and Bioengineering, Malet Place Engineering Building, University College London, London WC1E 6BT, UK. E-mail: mark.white{at}ucl.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods and materials
 Results
 Discussion
 Conclusion
 References
 
The purpose of this study was to adapt the hepatic perfusion index (HPI) methodology previously developed for MRI to derive 3D parametric maps of HPI, and to investigate apparent differences in HPI maps between a group of colorectal cancer patients and controls. To achieve this, a new and simpler approach to HPI calculation which does not require measurements from the aorta or portal vein is introduced, and assessed with large liver regions of interest (ROIs) in patients and controls. Several example HPI maps showing localized variation are then presented. The subject group consisted of 12 patients with known colorectal metastases, and 13 control subjects referred for routine contrast-enhanced spine imaging with no history of neoplastic disease. HPI was evaluated from serial T1 volume acquisitions acquired over the course of a Gd-DTPA bolus injection. Regions of abnormal perfusion were visible on the HPI maps derived for the patient group, manifested as areas of locally increased HPI extending around the visible margins of known metastases evident on the conventional contrast-enhanced images. This method for MR voxel-based parametric mapping of HPI has the potential to demonstrate regional variations in perfusion at the segmental and subsegmental level.


   Introduction
Top
 Abstract
 Introduction
Methods and materials
 Results
 Discussion
 Conclusion
 References
 
Liver metastases account for a significant proportion of the mortality associated with colorectal malignancy [1]. Approximately 25% of colorectal cancer patients are thought to have occult liver metastases at the time of surgical resection of the primary tumour [2]. These smaller lesions are known to respond favourably to treatment [3–5], but are largely undetectable with standard imaging techniques. Metastatic disease in the liver is, however, typically associated with an increase in the arterial component of perfusion. This study introduces a technique for mapping these perfusion abnormalities with MRI and evaluates it in a group of patients with known colorectal liver metastases. Small, undetected lesions are typically associated with early disease recurrence and failure of a surgical cure; so a method with the potential for both detection and anatomical localization of such disease could offer considerable benefits in the effective management of patients with colorectal liver metastases, as well as related applications in monitoring liver disease progression and treatment response.

Hepatic perfusion index (HPI) is defined as the ratio of hepatic arterial part to total (arterial plus portal pport) perfusion:


Formula 001

Using dynamic scintigraphy, HPI has been shown to be abnormally increased in 94% of patients with colorectal liver metastases [6]. Furthermore, of those patients who developed liver metastases within 3 years of their primary resection, 97% had an abnormal HPI at presentation [7], indicating that the observed changes in blood flow might constitute an effective marker for the likely presence of otherwise occult metastases. Studies of HPI measurement with dynamic CT and Doppler ultrasound have produced similar results [8–12]. Much of this work has by necessity been retrospective in nature, but prospective studies using this technique have demonstrated a high degree of sensitivity to occult metastases [13–15].

Two main methods have been proposed for deriving perfusion measurements and HPI from CT time–density information. Both calculate arterial perfusion part by dividing the peak gradient in the liver during arterial phase by the peak enhancement of the aorta [8, 9, 16, 17], but they differ in their approach to estimating portal perfusion pport, as illustrated in Figure 1Go. In the "indirect" approach [9, 18] shown in Figure 1aGo, pport is calculated as the peak gradient in the liver during portal phase gport divided by the peak arterial enhancement Iaorta:


Formula 002

Two refinements are made in the "direct" approach [8, 18], shown in Figure 1bGo: the arterial component of hepatic uptake is removed before measuring portal uptake gradient (by subtracting a scaled enhancement curve from a purely arterial organ such as the spleen), and perfusion is estimated by dividing this corrected portal gradient Formula by peak enhancement measured in the portal vein Ipv.


Formula 003

Both techniques have been reported to show differences in perfusion between control subjects and patients with malignancy [8, 9, 18], but the "direct" method is more physiologically appropriate and, when applied to dynamic CT, provides portal perfusion values in closer agreement to those derived using other techniques [18]. This method is also less subject to errors arising from arterial washout and recirculation.


Figure 1
View larger version (19K):
[in this window]
[in a new window]

  		Figure 1. Directvs indirect methods of hepatic perfusion index (HPI) estimation. The shaded area represents the liver enhancement curve, while the dotted and dashed lines depict the arterial and portal contributions, respectively. The liver curve is separated into arterial and portal phases according to the time of peak splenic enhancement, and the peak gradients are superimposed as heavy black lines. For both methods the hepatic arterial perfusion is derived from the peak gradient of the arterial phase of the liver curve. (a) The "indirect" method calculates the portal perfusion from the peak enhancement of the portal phase of the liver curve, while (b) the "direct" method first subtracts the arterial component and then calculates the portal venous perfusion from the peak gradient of the derived portal liver curve.

 
More recently, HPI methodology has been adapted to MRI [19, 20]. However, both of these studies relied on a region of interest (ROI) approach, which is not ideally suited for use as a screening technique as it does not provide information with regard to the anatomical distribution of perfusion abnormalities. Voxel-based parametric mapping of the HPI has the potential to demonstrate regional variations in perfusion at the segmental and subsegmental level, which can potentially be used to anatomically localize areas of abnormal perfusion that may be associated with metastatic spread. The purpose of this study was to adapt the HPI methodology previously developed for MRI to derive 3D parametric maps of the HPI, and to investigate apparent differences in the HPI maps between a group of colorectal cancer patients with known metastases and a group of control subjects. In order to achieve this, a new method of HPI mapping is introduced and first assessed over large ROIs in both subject groups.


   Methods and materials
Top
 Abstract
 Introduction
 Methods and materials
Results
 Discussion
 Conclusion
 References
 
Subject group
The patient group consisted of nine men and three women (aged 44–78 years), with primary colorectal cancer confirmed by biopsy or resection, and liver metastases evident on previous imaging (ultrasound, CT, or MRI). The control group consisted of eight men and five women (aged 21–77 years), taken from patients referred for routine contrast-enhanced spine imaging with no history of neoplastic disease; all controls also had normal liver function tests. This study was performed with the approval of the local research ethics committee, and informed consent was obtained from all subjects prior to MRI.

MRI protocol
Imaging was performed with a Siemens 1 T Harmony scanner (Siemens, Erlangen, Germany) with a surface array coil using a modified volume interpolated breath hold imaging (VIBE) sequence [21]. For each volume, 36 axial slices were obtained with a field of view of 400x300 mm and a matrix size of 128x67 (zero-filled to 128x96), resulting in an in-plane resolution of 3.1x3.1 mm (slices were 5 mm thick). A repetition time of 2 ms, an echo time of 0.83 ms, and a bandwidth of 1150 Hz were used to enable fast sampling. In total, 25 volumes (each consisting of 36 slices) were acquired with a temporal resolution of 3 s, resulting in a total scan time of 1 min 15 s. Since T2 saturation effects can lead to a non-linear relationship between signal intensity and contrast concentration, phantom experiments were performed with varying concentrations of gadolinium in saline and blood in order to optimize the VIBE protocol for a near-linear signal response to typical tissue contrast concentrations [19]. A flip angle of 90° was chosen to reduce the impact of B1 and B0 field variations on contrast weighting [19].

The VIBE sequence was acquired during gentle breathing in order to avoid flow variations introduced by changes in intra-abdominal pressure [22]; this was effective and few motion artefacts were seen within acquired image volumes. The patients were cannulated with a 22-gauge venflon in an antecubital vein, which was flushed with saline and connected to a MRI-compatible injector kit (Medex, Annecy-Le-Vieux, France). A bolus of the manufacturer-recommended dose of 0.2 ml kg–1 Gd-DTPA (Magnevist, Schering, Germany), typically 16 ml, was administered manually at a rate of 4 ml s–1 at the start of the VIBE sequence, and followed immediately by a 20 ml saline flush at the same injection rate. All subjects fasted for 3 h prior to the study.

Image processing and analysis
As noted above, the "direct" method of HPI estimation [8] is physiologically appropriate, and has previously been applied to dynamic MRI [19]. However, since this approach requires the measurement of peak aortic and portal venous enhancement, it is more complex and difficult to automate in a manner suitable for voxel-based HPI mapping. We therefore introduce a "combined" method, in which portal perfusion pport is derived from gradient Formula after subtraction of the arterial component from the liver curve (as in the direct method), but scaled to the enhancement of the aorta rather than that of the portal vein.


Formula 004

This approach makes the implicit assumption that a single blood concentration of contrast agent is applicable to both the arterial and portal perfusion, as in the "indirect" method used for dynamic CT measurements of the HPI [9]. However, since the relative hepatic and portal perfusion are both scaled to the peak of the aortic enhancement curve, this scaling factor appears in both the numerator and the denominator of the HPI (Equation (1)) and can be omitted entirely when only the HPI (and not individual calibrated perfusion estimates) are required. The "combined" method therefore lends itself to voxel-based HPI analyses because it removes the need for measurement of the aortic and portal venous concentrations of contrast agent. This enables the HPI analysis to be performed with minimal operator intervention, using rapid lower-resolution dynamic scans where the portal vein is difficult to isolate and where saturation effects in the aorta are hard to eliminate.

In order to establish the validity of this "combined" method for visualizing disease-related variations in HPI, all three methods (combined, direct and indirect) were first applied to large liver ROIs (excluding major blood vessels) as described by Totman et al [19]; these were drawn by a clinical specialist MRI radiographer and verified by a consultant radiologist. A two-tailed t-test was applied to determine whether significant differences existed between patients and controls in each case, and the t-statistics from the methods compared to indicate whether the "combined" method used for HPI mapping is able to separate normal from diseased tissue as well as the approaches it is derived from.

HPI analysis was implemented using the R statistical computing environment [23] and performed offline after DICOM export and anonymization of the data. Following the methodology established for CT measurements of the HPI [8, 9] and subsequently adapted for MRI by Totman et al [19], the shape of the splenic enhancement was used as an arterial reference. A ROI was drawn around the spleen (consensus between a physicist and a clinical specialist MRI radiographer), with margins slightly within the boundary of the spleen to minimize partial volume effects from other tissues during respiratory motion. The mean splenic uptake curve over this region was smoothed using a boxcar running mean; the time of peak splenic enhancement was used to divide the liver time series into arterial and portal phases (assuming that no portal enhancement takes place during the arterial phase, but that a component of arterial perfusion and washout continues during the portal phase).

To produce HPI maps, time-series data for each liver voxel were smoothed as for the spleen curve. In each voxel, the arterial gradient gart was taken to be the maximum liver gradient during arterial phase. The smoothed spleen curve was then scaled (again, separately at each point in the liver) such that its maximum arterial gradient matched gart of the liver voxel, and subtracted from the liver curve giving an estimated portal curve as described in [8]; this was used to calculate peak gradient of the voxel during portal phase, Formula . HPI was derived for each voxel as shown in Equations (1) and (4). Each HPI estimate in the map is therefore dependent only on the relative uptake gradients in the two phases at the single spatial position it is associated with, as well as on the shape and duration (but not the intensity scaling) of the spleen curve. Points in the data with identical tissue behaviour but different MRI intensity scaling, such as that caused by differing positions in the coil sensitivity profile, would give identical HPI values. Non-linear artefacts, such as local flip angle variations caused by B1 and B0 inhomogeneity, will still affect mapped HPI to some extent (and note that the large ROI-based calculations described above, unlike HPI maps, are not independent of linear intensity scaling). Generation of 3D HPI maps took a few minutes; a faster implementation would be possible if required for practical day-to-day use.


   Results
Top
 Abstract
 Introduction
 Methods and materials
 Results
Discussion
 Conclusion
 References
 
The distribution of HPIs calculated using the ROI implementation of the "combined" method is shown in Figure 2Go (boxes show median and interquartile ranges). Significant differences were seen between the perfusion indices calculated for the patient and control groups using the "combined" (p<0.005), "direct" (p<0.005) and "indirect" (p<0.01) methods (unpaired t-test), with patients demonstrating increased HPI relative to controls. The t-statistic calculated for the "combined" method (t = 3.9) was comparable with those derived for the "direct" method (t = 4.2) and "indirect" method (t = 3.3). The mean HPI values computed by each method for the patient and control groups are shown in Table 1Go.


Figure 2
View larger version (12K):
[in this window]
[in a new window]

  		Figure 2. Box-and-whisker plot showing overall hepatic perfusion index (HPI) calculated (using the "combined" method) from controls and from patients with confirmed primary colorectal cancer and metastatic liver disease evident on imaging. Points more than one and a half times the interquartile ranges from the median are plotted as outliers.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Mean hepatic perfusion index(HPI) values by the indirect, direct and combined methods over liver regions of interest

 
HPI maps were generated for all subjects. Because the HPI is only meaningful within the liver, for display purposes each HPI map is masked to the liver and overlaid on a corresponding post-contrast T1 weighted image (taken 12 s after peak splenic enhancement). An example HPI map from one of the control subjects is shown in Figure 3Go (right), with a post-contrast T1 weighted image shown in Figure 3Go (left) for comparison. The blood vessels in the HPI map are clear and well-differentiated, but the HPI in the surrounding tissue is fairly homogenous, typical of a disease-free liver.


Figure 3
View larger version (53K):
[in this window]
[in a new window]

  		Figure 3. Hepatic perfusion index(HPI) map from a control subject. (a) A T1 weighted image acquired just after the peak arterial enhancement, for comparison. (b) The same T1 image is shown in the background, overlaid with an HPI map of the liver. The HPI scale is shown to the right.

 
Figure 4Go shows the HPI map and corresponding post-contrast T1 weighted image from a patient with multiple confirmed metastases in the right liver, clearly visible as dark areas on the post-contrast T1 weighted images. A bright area is visible along the posterior border of the liver in the HPI map, corresponding to an area of abnormal arterialization. The extent of this abnormally perfused region is somewhat larger than the capsular metastases visible on the conventional contrast-enhanced images, potentially indicating an extended zone of angiogenesis or further metastatic pathology. These HPI increases correspond roughly to areas of early enhancement in the dynamic image series, but their differentiation is stronger on the HPI map due to the combined effects of increased arterial and reduced portal flow. These effects are depicted graphically in GoFigures 5 and 6Go, which show the arterial and portal phases of the liver curve derived from example voxels in the parenchyma and an area of arterialized tumour periphery, respectively (selected following discussion of the images with an expert radiologist). The maximum slope for each curve is indicated on the figure panels. As expected, the arterialized tumour voxel demonstrates significantly increased arterial perfusion and decreased portal perfusion relative to the voxel in the parenchyma.


Figure 4
View larger version (47K):
[in this window]
[in a new window]

  		Figure 4. Hepatic perfusion index map(and corresponding post-contrast T1 weighted image) from a patient with multiple confirmed metastases in the right liver. A bright area is visible along the posterior border of the liver (arrow) corresponding to an area of abnormal arterialization.

 

Figure 5
View larger version (11K):
[in this window]
[in a new window]

  		Figure 5. Perfusion estimation from enhancement gradient in a single parenchyma voxel. In each figure the dashed line shows the splenic arterial peak, separating arterial from portal phase; the red line shows the smoothed uptake curve from which gradients are taken; the black line shows measured gradient. (a) Peak uptake gradient during the arterial phase. (b) The estimated portal contribution (after subtraction of a scaled spleen curve), and peak gradient measured from this during the portal phase. The measured gradient during the portal phase in this parenchyma voxel exceeds that in the arterial phase.

 

Figure 6
View larger version (11K):
[in this window]
[in a new window]

  		Figure 6. Perfusion estimation in a single voxel in the tumour periphery(see Figure 5Go caption for explanation). The measured gradient during arterial phase is considerably higher than that in the portal phase.

 
Further examples of pathological HPI maps are shown in GoFigures 7 and 8Go. Figure 7Go shows the HPI map and corresponding post-contrast T1 weighted image for a patient with confirmed metastases in segments II and III of the left liver. A bright area is visible on the HPI map extending around the margin of the metastasis visible on the post-contrast T1 weighted image. Figure 8Go shows the HPI map and associated T1 weighted image for a patient with a large mass occupying much of segments VI–VIII of the right liver. The HPI map shows considerable heterogeneous arterialized structure within the large mass, including an area of increased HPI around the perimeter. The arrow indicates the small area of normal parenchyma; this area is bright on the post-contrast T1 weighted images, indicating that the total uptake of gadolinium is greater in this area than in the large mass. However, since the enhancement of the normal parenchyma takes place later (during the portal phase), the HPI map correctly shows this area as darker than that of the large mass, in keeping with a lower arterial component of perfusion.


Figure 7
View larger version (53K):
[in this window]
[in a new window]

  		Figure 7. Hepatic perfusion index(HPI) map (and corresponding post-contrast T1 weighted image) from a patient with confirmed metastases in segments II and III of the left liver. A bright area on the HPI map extends around the margins of the metastasis visible on the T1 weighted post-contrast image (arrow).

 

Figure 8
View larger version (49K):
[in this window]
[in a new window]

  		Figure 8. Hepatic perfusion index(HPI) map (and corresponding post-contrast T1 weighted image) from a patient with a large mass occupying much of segments VI–VIII of the right liver. Considerable heterogeneous structure is visible within the mass in the HPI map, including increased HPI around the perimeter. An area of normal-appearing parenchyma is also visible (arrow). Although this area is bright in post-contrast T1, indicating greater total gadolinium uptake than in the large mass, enhancement occurs mainly during the portal phase, and the corresponding HPI is correctly shown as low.

 

   Discussion
Top
 Abstract
 Introduction
 Methods and materials
 Results
 Discussion
Conclusion
 References
 
Current imaging techniques can reliably demonstrate metastases of 1 cm or larger using MRI and multidetector CT, but the sensitivity of imaging methods to smaller lesions is in the region of 50% [24]. Since it is these undetected, sub-centimetre lesions that typically cause early disease recurrence resulting in failure of a surgical cure, a method capable of early detection of these smaller lesions could offer substantial benefits in the management of patients with suspected colorectal liver metastases. Such a method could conceivably facilitate the early identification of patients suitable for resection, cryoablation, or chemotherapy, and improve the efficacy of these treatments with small volume disease.

The HPI has been used with several imaging modalities to characterize the changes in perfusion associated with metastatic disease in the liver [6–11, 19], but there are several apparent advantages to a MR-based approach. The equivalent marker index developed for Doppler ultrasound (the Doppler perfusion index, or DPI) is based on blood flow in feeding vessels, and does not provide any information regarding the anatomical distribution of hepatic perfusion. It relies upon "normal" hepatic arterial anatomy, but only 70% of the population demonstrate such anatomy [25]. DPI data is also subject to significant interobserver and intraobserver variability [25]. In some cases, radiation exposure in CT and radionuclide studies (but avoided with MRI) may be a concern. Although advances in multislice CT may now allow improved coverage, to date, the CT HPI has only been demonstrated on a single slice and the iodinated contrast agents used in CT carry an attendant risk of anaphylaxis. In comparison, gadolinium is considered a safer contrast agent and the volume of the injected bolus is smaller, which facilitates data processing and reduces errors potentially arising from washout of the contrast agent before the peak gradient of the portal liver curve. Unlike ultrasound DPI, this MR-based technique can be used with abnormal hepatic arterial anatomy, and it provides increased spatial and temporal resolution in comparison with radionuclide based methods. In addition, this technique offers the potential for assessment of the whole liver volume as part of a routine MR examination including high-resolution anatomical imaging. Although dynamic MRI protocols inevitably involve balancing spatial and temporal resolution requirements, the VIBE sequence used in this work demonstrates both adequate spatial resolution and coverage for regional HPI visualization, and a temporal resolution giving sufficient time points to quantify uptake gradients. The non-linear relationship between MR signal intensity and contrast concentration can confound the quantification of perfusion in absolute units, but the HPI ratio calculated from MR data has yielded results consistent with those from previous dynamic CT, Doppler ultrasound, and scintigraphy studies and appears to represent a valid physiological indicator of abnormal perfusion [19].

The extension of the MRI HPI technique to produce parametric maps of the HPI can enable the visualization of areas of abnormal perfusion, which may be associated with metastatic disease progression. This approach can be used to complement and guide ROI-based measurements of the HPI, and could potentially facilitate the early detection of occult lesions. The group mean HPI values yielded by the "combined" method are close to those of the "indirect" method, following from the shared assumption of identical hepatic arterial and portal venous blood contrast concentration. These absolute HPI values are higher than those given by the "direct" method previously implemented for MRI [19], which estimates portal venous concentration from the data. However, the comparable t-statistics calculated for the "combined" and "direct" techniques indicate that these methods are equally effective in detecting differences in perfusion between colorectal cancer patients and controls, and are arguably equally valid for the purposes of visualization of HPI differences in parametric maps. In addition, since the proposed "combined" technique does not require measurement of the aortic or portal venous concentrations of contrast agent, the errors associated with interoperator variability and partial volume effects are significantly reduced. Saturation and flow effects will influence signal intensity in the high aortic peak contrast concentrations in the data presented here, leading to slight (although relatively consistent) errors in HPI values estimated by the "direct" method; the "combined" method is free from these errors. The mapping approach presented also yields voxel-based HPI values largely unaffected by scaling artefacts such as the receiver sensitivity profile.

Nevertheless, further work will be required to fully assess the usefulness and of reproducibility of this technique. Some overlap was shown between the overall HPI of controls and patients (Figure 2Go), and while the separation is sufficient to suggest that patients demonstrating HPI elevated into the overlap region may be at higher risk and should thus undergo chronologically closer follow-up imaging, a larger subject group would allow better estimation of a threshold for "abnormal" HPI in individual patients. The presence of pre-existing liver disease such as cirrhosis, other liver tumours, or chemotherapy effects could potentially induce alterations in perfusion unrelated to the presence of metastases, and follow-up HPI imaging of a large cohort of patients could clarify both the specificity of HPI changes and the value of HPI imaging in the management of patients with colorectal cancer. Although some inferences about localized HPI changes can be drawn from known correspondences between elevated HPI and disease progression [15, 26], the exact physiological meaning of spatial variations in HPI is still unknown; a prospective study including HPI and appropriate histopathological assessment, or work in animal models, would be needed to increase understanding of the relationship between HPI changes and occult metastases. The use of gentle breathing during dynamic imaging, while not presenting a problem for ROI HPI analysis, did cause local errors in HPI (blurring of HPI maps) in a small number of cases due differences in position of small structures between adjacent time points, especially around the time of contrast arrival; another area for further work would be the use of image registration to reduce these artefacts and increase effective mapping resolution. Finally, this study did not take into account age and gender related differences in hepatic perfusion; future studies will consider the impact of these factors.


   Conclusion
Top
 Abstract
 Introduction
 Methods and materials
 Results
 Discussion
 Conclusion
References
 
HPI mapping with dynamic gadolinium-enhanced MRI can potentially characterize both the spatial extent and anatomical distribution of changes in perfusion associated with metastatic disease in the liver. This technique may provide a valuable method for early detection of occult liver metastases, for monitoring disease progression, and for assessing localized changes in hepatic perfusion during therapy.

Received for publication September 12, 2005. Revision received April 28, 2006. Accepted for publication May 30, 2006.


   References
Top
 Abstract
 Introduction
 Methods and materials
 Results
 Discussion
 Conclusion
 References
 

  1. Abrahams HL, Spiro R, Goldstein N. Metastases in carcinoma. Cancer 1950;3:74–85.[CrossRef][Medline]
  2. Mainprize KS, Mortensen NJ, Warren BF. Early colorectal cancer: recognition, classification, and treatment. Br J Surg 1998;85:469–76.[CrossRef][Medline]
  3. Moertel CG, Fleming TR, Macdonald JS, Haller DG, Laurie JA, Goodman PJ, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 1990;322:352–8.[Abstract]
  4. Laurie JA, Moertel CG, Fleming TR, Wieand HS, Leigh JE, Rubin K, et al. Surgical adjuvant therapy of large-bowel carcinoma: an evaluation of levamisole and the combination of levamisole and fluorouracil. J Clin Oncol 1989;7:1447–56.[Abstract]
  5. Taylor I, Machin D, Mullee M, Trotter G, Cooke T, West C. A randomized controlled trial of adjuvant portal vein cytotoxic perfusion in colorectal cancer. Br J Surg 1985;72:359–63.[Medline]
  6. Breedis C, Young G. The blood supply of neoplasms in the liver. Am J Pathol 1954;30:969–85.[Medline]
  7. Cooke DA, Parkin A, Wiggins P, Robinson PJ, Giles GR. Hepatic perfusion index and the evolution of liver metastases. Nucl Med Commun 1987;8:970–4.[Medline]
  8. Blomley MJ, Coulden R, Dawson P, Kormano M, Donlan P, Bufkin C, et al. Liver perfusion studied with Ultrafast CT. J Comput Assisted Tomogr 1995;19:424–33.[Medline]
  9. Miles KA, Hayball MP, Dixon AK. Functional images of hepatic perfusion obtained with dynamic CT. Radiology 1993;188:405–11.[Abstract/Free Full Text]
  10. Leen E, Angerson WG, Cooke TG, McArdle CS. Prognostic power of Doppler perfusion index in colorectal cancer. Correlation with survival. Ann Surg 1996;223:199–203.[CrossRef][Medline]
  11. Leggett DA, Kelley BB, Bunce IH, Miles KA. Colorectal cancer: diagnostic potential of CT measurements of hepatic perfusion and implications for contrast enhancement protocols. Radiology 1997;205:716–20.[Abstract/Free Full Text]
  12. Harvey CJ, Blomley MJ, Eckersley RJ, Cosgrove DO. Developments in ultrasound media. Eur Radiol 2001;11:675–89.[CrossRef][Medline]
  13. Cuenod CA, Leconte I, Siauve N, Resten A, Dromain C, Poulet B, et al. Early changes in liver perfusion caused by occult metastases in rats: detection with quantitative CT. Radiology 2001;218:556–61.[Abstract/Free Full Text]
  14. Nott DM, Grime SJ, Yates J, Day DW, Baxter JN, Jenkins SA, et al. Changes in the hepatic perfusion index during the development of experimental hepatic tumours. Br J Surg 1989;76:259–63.[Medline]
  15. Warren HW, Gallagher H, Hemingway DM, Angerson WJ, Bessent RG, Wotherspoon H, et al. Prospective assessment of the hepatic perfusion index in patients with colorectal cancer. Br J Surg 1998;85:1708–12.[CrossRef][Medline]
  16. Peters AM, Gunasekera RD, Henderson BL, Brown J, Lavender JP, De Souza M, et al. Noninvasive measurement of blood flow and extraction fraction. Nuc Med Commun 1987;8:823–37.
  17. Peters AM, Brown J, Hartnell GG, Myers M, Haskell MJ, Lavender JP. Non-invasive measurement of renal blood flow with Tc-99m DTPA: comparison with radiolabelled microspheres. Cardiovasc Res 1987;21:830–4.[Medline]
  18. Tsuchima Y, Blomley MJK, Kusanu S, Endo K. Measuring portal venous perfusion with contrast-enhanced CT: comparison of direct and indirect methods. Acad Radiol 2002;9:276–82.[CrossRef][Medline]
  19. Totman JJ, O'Gorman RL, Kane PA, Karani JB. Comparison of the hepatic perfusion index measured with gadolinium-enhanced volumetric MRI in controls and in patients with colorectal cancer. Br J Radiol 2005;78:105–9.[Abstract/Free Full Text]
  20. White MJ, Koh DM, Moore L, Collins DJ, Leach MO. Assessing reproducibility of HPI measurements in rapidly-enhancing liver metastases using fast central k-space MR acquisition. MAGMA (ESMRMB) 2003;16:S294
  21. Lee VS, Lavelle MT, Rofsky NM, Laub G, Thomasson DM, Krinsky GA, et al. Hepatic MR imaging with dynamic contrast enhanced isotropic volumetric breath hold examination: feasibility, reproducibility and technical quality. Radiology 2000;215:365–72.[Abstract/Free Full Text]
  22. Sadek AG, Mohamed FB, Outwater EK, El-Essawy SS, Mitchel DG. Respiratory and postprandial changes in portal flow rate: assessment of phase contrast MR imaging. J Magn Reson Imaging 1996;1:90–3.
  23. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2005
  24. Ward J, Naik KS, Guthrie JA, Wilson D, Robinson PJ. Hepatic lesion detection: comparison of MR imaging after the administration of superparamagnetic iron oxide with dual-phase CT by using alternative-free response receiver operating characteristic analysis. Radiology 1999;210:459–66.[Abstract/Free Full Text]
  25. Fowler RC, Harris KM, Swift SE, Ward M, Greenwood DC. Hepatic Doppler perfusion index: measurement in nine healthy volunteers. Radiology 1998;209:867–71.[Abstract/Free Full Text]
  26. Ballantyne KC, Charnley RM, Perkins AC, Pye G, Whalley DR, Wastie ML, et al. Hepatic perfusion index in the diagnosis of over metastatic colorectal cancer. Nucl Med Commun 1990;11:23–8.[Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
K. W. Kim, J. M. Lee, E. Klotz, H. S. Park, D. H. Lee, J. Y. Kim, S. J. Kim, S. H. Kim, J. Y. Lee, J. K. Han, et al.
Quantitative CT Color Mapping of the Arterial Enhancement Fraction of the Liver to Detect Hepatocellular Carcinoma
Radiology, February 1, 2009; 250(2): 425 - 4s34.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by White, M J
Right arrow Articles by Totman, J J
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by White, M J
Right arrow Articles by Totman, J J

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS