| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Institute for Diagnostic and Interventional Radiology (C.H., T.D., T.J.V.) and Departments of Thoracic and Cardiovascular Surgery (S.D., M.F.K., G.W.G., A.M.) and Epidemiology and Medical Statistics (H.A.), J. W. Goethe-University Frankfurt, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany; and CT Division, Siemens Medical, Forchheim, Germany (S.S., T.G.F.). Received May 29, 2002; revision requested July 26; final revision received December 17; accepted January 14, 2003. Address correspondence to C.H. (e-mail: c.herzog@em.uni-frankfurt.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-six patients preoperatively underwent multi–detector row CT (4 x 1-mm collimation, pitch of 1.5, 500-msec rotation time, retrospective electrocardiographic gating, 1.25-mm effective section thickness) and coronary angiography. Assessment criteria for both techniques were visibility and cardiac course of coronary arteries, localization and degree of stenoses, composition of atherosclerotic plaques, and vascular diameter at anastomosis site. Site for distal bypass anastomosis was recommended. Results at multi–detector row CT were calculated relative to results at coronary angiography and surgery.
RESULTS: Multi–detector row CT properly displayed 79.4% (154 of 194) of all surgical relevant coronary segments and 80.4% (434 of 540) of all coronary segments. For coronary angiography, ratios of 88.7% (172 of 194) and 94.6% (511 of 540), respectively, were observed. For detection of calcified plaques, multi–detector row CT results exceeded those at coronary angiography by a difference of 17% (18 of 18 [100%] compared with 15 of 18 [83%]). Hemodynamically relevant stenoses were identified with multi–detector row CT in 76% (42 of 55) of cases. Bridging of coronary segments through either myocardium (four of five) or epicardial fat (two of three) was better identified at multi–detector row CT than it was at coronary angiography (one of five compared with zero of three, respectively). At multi–detector row CT, 76% (28 of 37) of all distal bypass touchdown segments were identified, but at coronary angiography, only 70% (26 of 37) were identified.
CONCLUSION: Multi–detector row CT provides extended information about coronary target site and therefore should be regarded as an ideal additive planning tool for complex minimally invasive procedures such as totally endoscopic coronary artery bypass grafting or minimally invasive direct coronary artery bypass grafting.
© RSNA, 2003
Index terms: Computed tomography (CT), multi–detector row, 54.12119 • Computed tomography (CT), three-dimensional, 54.12117 • Coronary angiography, comparative studies, 54.124, 54.1244 • Coronary vessels, bypass graft, 54.126
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
However, one of the technical drawbacks in TECAB procedures is the lack of tactile feedback. For the dissection and exposure of intramural coronary arteries or vessels hidden deep inside the epicardial fatty tissue, one relies mainly on visual information. Thus, performance of this procedure with TECAB is technically much more challenging than it is with conventional surgery (4). Preoperative coronary angiography unfortunately often does not provide all relevant visual information for these minimally invasive surgical procedures. Thus, if collateral circulation is absent, the coronary territory distal to a total vascular occlusion often cannot be displayed properly. Consequently, the problem with displaying the coronary territory renders impossible any sufficient evaluation of wall quality, plaque composition, and vascular diameter of the target anastomotic site. In addition, even if visualization of the diseased vessel and its distal segments is possible, angiograms often lack other important morphologic information, such as the relative position of the vessel in relation to the surrounding cardiac fatty tissue or its exact cardiac course (5).
The aim of this study, therefore, was to assess the use of multi–detector row spiral computed tomography (CT) in the preoperative evaluation of the target site in patients who were undergoing the TECAB procedure performed on the arrested or beating heart and to correlate the data with coronary angiographic and intraoperative findings.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
CT Imaging
All CT scans were obtained by using a multi–detector row spiral CT scanner (Somatom Plus 4 VolumeZoom WIP-version VA 20; Siemens, Forchheim, Germany). Patients with heart rates greater than 70 beats per minute previously received a short-lasting beta-blocker (esmolol hydrochloride, Brevibloc; Baxter, Unterschleiáheim, Germany) administered as 10 mg per 10 kg of body weight to obtain rates of 60 beats per minute or less. Each patient first underwent contrast material–enhanced and, subsequently, nonenhanced examinations. Scanning parameters were 120 kV and 300 mAs, 500-msec rotation time, 4 x 2.5-mm section collimation, and 3.8-mm table feed per rotation for the nonenhanced series and 4 x 1.0-mm section collimation and 1.5-mm table feed per rotation for the contrast-enhanced series. All patients received 140–160 mL of a nonionic contrast medium (Ultravist; Schering, Berlin, Germany) infused through an 18-gauge intravenous antecubital catheter at a flow rate of 3.5 mL/sec. Start delay was calculated with a test bolus placed in the ascending aorta; 30 mL of contrast medium was administered at a flow rate of 3.5 mL/sec.
Image Reconstruction
Image reconstruction was performed by using retrospective electrocardiographic gating, a technique that allowed continuous image reconstruction from volume data sets during any phase of the cardiac cycle (6,7). Reconstruction parameters were a 220-mm field of view; kernel B35, a medium soft-tissue kernel; 3.00-mm effective section thickness; and a 1.50-mm increment for the nonenhanced series; those for the contrast-enhanced series were a 1.25-mm effective section thickness and a 0.5-mm increment.
For image reconstruction, the adaptive cardiac volume reconstruction algorithm, which is standardized and provided with the software of the spiral CT scanner, was used. Depending on the heart rate, this algorithm combined two different reconstruction procedures: single-segmental reconstruction (
65 beats per minute) and adaptive-segmental reconstruction (>65 beats per minute). Single-segmental reconstruction requires data from only one rotation; adaptive-segmental reconstruction requires data from at least two rotations for the reconstruction of each 1.25-mm section (8). Although temporal resolution for single-segmental reconstruction constantly was 250 msec, that for adaptive-segmental reconstruction that was dependent on the heart rate ranged between 250 msec at maximum and 125 msec at minimum (8). Taking into account basic cardiac physiology, sufficient image reconstruction seemed to be feasible solely if performed between the late systole (ie, ascending T wave) and the late diastole (ie, beginning of the P wave) (9). Each data set was consequently reconstructed at multiple time points within this interval, and the time points differed from each other by 50 msec. Because of a decrease in the length of the TP interval at increasing heart rates, the number of possible reconstruction time points per patient ranged between 10 (low heart rate) and six (elevated heart rate). At each time point, image reconstruction was performed antegrade and absolute (ie, in milliseconds) in relation to the R peak. Subsequently, for each patient and each main coronary artery (right coronary artery, LCA, left circumflex artery) separately, one specific reconstruction was determined that showed the fewest motion artifacts and therefore made possible a proper image interpretation (10–14).
Image Reformation and Evaluation
Multi–detector row CT image evaluation was performed by using both transverse scans and secondary reformations, that is, multiplanar and three-dimensional reformations. Transverse scans and multiplanar reformations were analyzed at one workstation (VZ Wizzard; Siemens, Erlangen, Germany), and three-dimensional reformations were analyzed at another workstation (3D Virtuoso; Siemens) by using a volume-rendering technique. Transverse scans and multiplanar reconstructions were displayed on a 512 x 512 matrix, and three-dimensional reformations were displayed on a 256 x 256 matrix. After initial blinding regarding the data set, all data sets were assessed by two independent observers (T.D., C.H.) separately.
Criteria analyzed were as follows: the total visibility of coronary artery segments and the visibility of those segments that were of surgical relevance, the atherosclerotic plaque load of each single coronary artery segment, the composition of the individual plaques, and the epicardial or intramural course of the vessel. Plaque load and plaque composition were determined solely on transverse scans and multiplanar reformations. Visibility and cardiac course of the coronary vessels were additionally assessed on three-dimensional reformations, which also served for surgical morning round demonstrations and intraoperative orientation. Visibility was determined by the ability of the reader to distinguish coronary artery segments as visible or not visible.
The segmental plaque load was assessed according to the classification of the American Heart Association (AHA). According to the AHA classification, the coronary artery territory is subcategorized into 15 segments, and six degrees of atherosclerosis are distinguished (15): type I, irregular wall outline with less than 25% stenosis; type II, slight stenosis (25%–50%); type III, moderate stenosis (51%–74%); type IV, hemodynamically relevant stenosis (75%–89%); type V, subtotal stenosis (90%–99%); and type VI, vascular occlusion (100%). The prevailing degree of occlusion was ascertained with an automated distance measuring tool (Plus4VZ Wizzard; Siemens).
In regard to the composition of the plaque, a distinction was made between calcified and noncalcified plaques. Plaques with a mean attenuation of 130 HU or greater were graded as calcified, whereas plaques with a mean attenuation of less than 130 HU were graded as noncalcified (16). Calcified plaques were identified on nonenhanced scans, and noncalcified plaques were identified on contrast-enhanced scans.
The cardiac course of each coronary artery was assessed by distinguishing a possible epicardial or intramural course of the prevailing arterial segment, according to specifications of the AHA (1–15).
Finally, with consideration of the results of all three reformation techniques, a conclusive recommendation was made in regard to which coronary segment was best suited for allocation of the distal bypass anastomosis, and the coronary diameter in this region was calculated.
All findings were compared with those on corresponding coronary angiograms, which had been obtained with different technical systems and the Judkins technique. At least four views of the left and two views of the right coronary artery system were analyzed in consensus by two observers (S.D., M.F.K.) who were both trained in this technique. To avoid recall bias, neither of them had any knowledge of the CT results, and evaluation was undertaken only on data sets that had previously been blinded.
Plaque composition was described as either calcified or noncalcified, and coronary artery segments were described as either visible or nonvisible. A search was conducted for calcified plaques on both nonenhanced and contrast-enhanced series.
The prevailing degree of stenosis and the segmental diameter at the planned site of distal anastomosis were determined by using a stenosis grading tool with automatic distance and scale calibration (Osiris; University Hospital Geneva, Switzerland).
As with multi–detector row CT, a conclusive recommendation was made in regard to the optimal place for distal bypass anastomosis, and the segmental diameter in this region was calculated.
All results were intraoperatively evaluated by one of the surgeons (S.D.) by using three-dimensional reformations and angiograms for comparison. Because of a restricted field of vision, only those segments that could be visualized through the surgical microscope were compared. These segments included the stenotic or occluded segment itself, the segments before and behind this occlusion or stenosis, and any surrounding side branches and possible crossing veins. All differences, as well as conformities, were noted and additionally documented on videotape; this allowed a second inspection after the operation.
Surgical Technique for TECAB Procedures
In a typical left-sided approach for left internal mammary artery to LCA bypass grafting, the patient is placed on the operating table in a supine position, with the left side of the chest elevated 30°–40°. Usually the fifth intercostal space close to the anterior axillary line is identified, and after deflation of the left lung, a camera port is placed bluntly to avoid left ventricular injury. The chest is insufflated with warm (37°C) CO2. After insertion of the endoscope, two ports are placed within visual control to accommodate the two robot arms, usually in the third and seventh intercostal space (Fig 1). The left internal mammary artery is mobilized from the subclavian artery down to the distal bifurcation by using an endoscope angled 30° upward. The distal end of the left internal mammary artery is skeletonized for grafting, and a soft bulldog clamp (Scanlan, St Paul, Minn) is placed (1,3).
|
|
statistic. The results were interpreted according to the following scale: poor, 0.20 or less; fair, 0.21–0.40; moderate, 0.41–0.60; good, 0.61–0.80; very good, 0.81–0.90; or excellent, 0.91–1.00. A 95% CI, calculated with a standard method, was assigned to the calculated value. Possible differences between both methods were tested for significance by using a comparison of Poisson frequencies.
In regard to the detection and grading of atherosclerosis and with reference to all 15 AHA segments, a coefficient of agreement for both techniques was calculated by means of a binominal CI for 
. The agreement between the methods and investigators was interpreted according to the and values, respectively, as poor (0.20), fair (0.21–0.40), moderate (0.41–0.60), good (0.61–0.80), very good (0.81–0.90), or excellent (0.91–1.00). A 95% CI, calculated with a standard method, was assigned to each calculated and value.
Sensitivity, specificity, and positive and negative predictive values of multi–detector row CT and coronary angiography in the detection of hemodynamically relevant stenoses were determined with subclassification of AHA atherosclerosis types I-VI into two different groups: patients without hemodynamically relevant stenoses (ie, <75%, AHA types I–III) and patients with hemodynamically relevant stenoses (ie, 
75%, AHA types IV–VI). Values for multi–detector row CT were calculated relative to all 15 AHA segments, with coronary angiography as the reference standard, as well as relative to those segments that were intraoperatively explored, with surgery as the reference standard. Values for invasive angiograms were calculated with only surgery as the reference standard.
Sensitivity, specificity, and positive and negative predictive values in the detection of calcified plaques were calculated for multi–detector row CT and coronary angiography relative to the surgical findings. The agreement between investigators for this evaluation was assessed by means of the statistic, as described previously. A 95% CI, calculated by using a standard method, was assigned to the calculated value.
The validity of both imaging methods in the correct identification of a segment’s cardiac course—on the epicardial surface, deep within the fatty tissue or within the myocardium—was calculated as proportional to the actual surgical findings.
The correspondence rate of multi–detector row CT and coronary angiographic findings in regard to the allocation of the distal bypass anastomosis and the segmental diameter at the site of anastomosis was also calculated proportional to the surgical results. A segmental diameter was regarded of equal size if the measured difference amounted to less than 1 mm. Bypass allocation was rated as correct if segmental allocation corresponded to that at surgery. Results concerning the cardiac course of the segment, the allocation of bypass anastomosis, and the segmental diameter at the site of anastomosis were obtained by both observers in consensus.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
value between investigators was 0.870 (95% CI: 0.786, 0.954), which indicated very good agreement. On average, 100% (108 of 108) of all proximal segments (AHA segments 1, 5, and 11), 94% (68 of 72) of the medial segments of the right coronary artery (AHA segments 2 and 3), 90% (65 of 72) of the medial segments of the LCA (AHA segments 6 and 7), 83% (30 of 36) of the medial segments of the left circumflex artery (AHA segment 13), 92% (33 of 36) of the distal segments of the right coronary artery (AHA segment 4), 81% (29 of 36) of the distal segments of the LCA (AHA segment 8), and 56% (20 of 36) of the distal segments of the left circumflex artery (AHA segment 15) were classified as visible. Side branches of the LCA (AHA segments 9 and 10) were regarded as visible in 71% (51 of 72); those of the left circumflex artery (AHA segments 12 and 14) were regarded as visible in 42% (30 of 72) (Table 1).
|
With consideration of only 194 of 540 segments that were intraoperatively explored on multi–detector row CT scans, 79.4% (154 of 194) could be evaluated by observer 1 and 76.8% (149 of 194) could be evaluated by observer 2. The value between investigators was 0.925 (95% CI: 0.784, 1.000), which represented excellent agreement. On angiograms, 88.7% (172 of 194) were visible to the observer (Table 1). With both methods, no significant differences in overall visualization were found for observer 1 (P = .346) or for observer 2 (P = .219).
The coefficient of agreement () for both techniques regarding detection and grading of atherosclerosis (AHA I-VI) was 0.759 (95% CI: 0.721, 0.795) for observer 1 and 0.731 (95% CI: 0.691, 0.768) for observer 2, which indicated good agreement for both. On average, at multi–detector row CT, 17.6% (29 of 165) of all stenoses were overestimated, and 6.7% (11 of 165) of them were underestimated.
When we considered all 15 AHA segments, sensitivity and specificity for multi–detector row CT and observer 1 in the identification of hemodynamically relevant stenoses (ie, those >75%) were 76% (42 of 55; 95% CI: 62.9%, 86.7%) and 99.6% (483 of 485; 95% CI: 98.5%, 99.9%), respectively. For observer 2, a sensitivity of 71% (39 of 55; 95% CI: 57.1%, 82.4%) and a specificity of 98.4% (477 of 485; 95% CI: 96.8%, 99.3%) were obtained. The positive predictive value was 95% (42 of 44; 95% CI: 84.5%, 99.4%) for observer 1 and 83% (39 of 47; 95% CI: 69.2%, 92.4%) for observer 2. The negative predictive value was 94.2% (467 of 496; 95% CI: 95.6%, 98.6%) for observer 1 and 96.8% (477 of 493; 95% CI: 94.8%, 98.1%) for observer 2 (Table 2). The value between investigators was 0.952 (95% CI: 1.000, 0.867), which was equivalent to excellent agreement.
|
were not calculated because of a lack of negative controls.
Both observers detected calcified plaques on multi–detector row CT scans, with 100% (18 of 18; 95% CI: 84.7%, 100%) sensitivity. For coronary angiograms, sensitivity was 83% (15 of 18; 95% CI: 58.6%, 96.4%) (Fig 3). Specificity was 100% (19 of 19; 95% CI: 85.4%, 100%) for both techniques and both multi–detector row CT observers. For multi–detector row CT, the positive predictive value was 100% (18 of 18; 95% CI: 84.7%, 100%), and the negative predictive value was 100% (19 of 19; 95% CI: 85.4%, 100%). For coronary angiography, the positive predictive value was 100% (15 of 15; 95% CI: 81.9%, 100%), and the negative predictive value was 86.4% (19 of 22; 95% CI: 65.1%, 97.1%) (Table 2). The value between investigators at multi–detector row CT was 1.00, which indicated excellent agreement.
|
|
|
Bridging of coronary artery segments through either myocardium or epicardial fatty tissue thus was identified more reliably on multi–detector row CT scans than on angiograms (Fig 5): 80% (four of five) compared with 20% (one of five) for myocardial bridging and 67% (two of three) compared with 0% (zero of three), respectively, for bridging through epicardial fatty tissue (Table 3).
|
On multi–detector row CT scans, seven segments were not assessed properly: In three, this occurred because of motion artifacts; in three, because of missing contrast media enhancement caused by a proximal total vascular occlusion; and in one, because of underestimation of the actual diameter.
On angiograms, all five segments that were not visualized adequately were located behind a total vascular occlusion.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Our results indicate that multi–detector row CT often can be used to identify such morphologic traps and therefore could improve the success rate of TECAB procedures. First, the selection of patients is facilitated, thus rendering possible preoperative switching to alternative surgical techniques. Second, virtual evaluation of the operative site by using three-dimensional reformations before surgery allows exact identification of the target vessel, its morphologic appearance, and its relationship to surrounding structures. The surgeon may preoperatively simulate the surgical approach and the exploration of the vessel. Although our results indicate that invasive coronary angiography could be used to more sensitively identify and display vascular occlusions, at multi–detector row CT, several morphologic traps that were not visualized with this technique could be detected.
Thus, findings at multi–detector row CT in four patients indicated that the target vessel was hidden deep inside the epicardial fat and those in two other patients indicated that multi–detector row CT showed an intramural course of the target vessel and therefore clearly outperformed invasive coronary angiography. With the latter technique, the location and course of the vessel were identified in only one patient and no patients, respectively. In four of these patients, the correct distal bypass touchdown segment was predicted correctly only at multi–detector row CT, whereas it was not predicted correctly at coronary angiography. In two other patients, only three-dimensional multi–detector row CT reformations showed the fatal combination of a prominent superficial diagonal branch and an intramural LCA. In such cases, accidental grafting of the wrong vessel (ie, diagonal branch versus LCA) is a common risk of endoscopic surgery because spatial orientation is difficult, the field of view is restricted, and manual palpation of the vessel is not possible (Fig 4).
Moreover, the composition of atherosclerotic plaques often cannot be predicted sufficiently on angiograms (5). Thus, calcified plaques were detected with a marked lower sensitivity on angiograms than they were on multi–detector row CT scans; with the multi–detector row CT scans, these plaques were identified in 100% of cases. Angiograms only display the lumen of the vessel and not the wall; thus, in case of a proximal total occlusion and at the same time an absence of collateral circulation, distal parts of this particular vessel often are not displayed accurately because no contrast medium reaches these parts. If collateral circulation is present, the stenosis is bypassed and the lumen is filled from the distal parts and, thus, the vessel can be seen. With multi–detector row CT, calcified plaques may at least be displayed because this modality offers the possibility of three-dimensional imaging (18–23). The results indicated previously support this assertion. However, such calcifications very often render successful bypass grafting in the region of choice impossible and thus either prolong operation times or imply intraoperative switching to alternative surgical procedures.
However, because of several profound restrictions, multi–detector row CT must still be regarded solely as a complementary but not as an alternate technique to invasive angiography. Because of its low temporal (125 msec) and spatial resolution, it is highly susceptible to motion artifacts—particularly at elevated heart rates—and offers only poor visualization of small coronary artery segments (24). The poor visibility obtained in the present study, particularly for the distal and marginal segments of the left circumflex artery, may thus be caused by the awkward combination of close proximity to the atrium and overall small vascular diameter. Other comparable coronary artery segments either were larger in diameter (eg, right coronary artery) or showed a more favorable localization (eg, diagonal and distal branches of the LCA).
In conclusion, the combination of the advantages of invasive coronary angiography (eg, high temporal and spatial resolution, blood flow information, assessment of functional parameters and collateral circulation) with the gross morphologic superiority of multi–detector row CT can be of great benefit to the operative outcome. Multi–detector row CT provides extended information about the coronary target site and therefore should be regarded as an ideal additive planning tool for complex minimally invasive procedures such as TECAB grating or minimally invasive direct coronary artery bypass grafting.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, T.J.V., A.M.; study concepts, C.H., S.D.; study design, C.H., T.J.V., T.D.; literature research, M.F.K., C.H., T.D.; clinical studies, C.H., S.D., G.W.G.; data acquisition, C.H., S.D., G.W.G.; data analysis/interpretation, C.H., S.D., M.F.K., H.A.; statistical analysis, H.A.; manuscript preparation, C.H., M.F.K.; manuscript definition of intellectual content, T.D., S.S., T.G.F., C.H.; manuscript editing, T.G.F., C.H.; manuscript revision/review, S.S., C.H., T.J.V.; manuscript final version approval, T.J.V., T.D., G.W.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. J. Rybicki, T. Sheth, and F. Y. Chen Cardiac Surgical Imaging Card. Surg. Adult, January 1, 2008; 3(2008): 179 - 198. [Full Text]
|
|
|
|
 |
|
 C. Herzog, S. A. Nguyen, G. Savino, P. L. Zwerner, J. Doll, C. D. Nielsen, T. G. Flohr, T. J. Vogl, P. Costello, and U. J. Schoepf Does Two-Segment Image Reconstruction at 64-Section CT Coronary Angiography Improve Image Quality and Diagnostic Accuracy? Radiology, July 1, 2007; 244(1): 121 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Stein, J. M. Smith, J. R. Robinson, and M. R. Katz Target vessel detection and coronary anastomosis assessment by intraoperative 12-MHz ultrasound. Ann. Thorac. Surg., September 1, 2006; 82(3): 1078 - 1084. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Herzog, M. Arning-Erb, S. Zangos, K. Eichler, R. Hammerstingl, S. Dogan, H. Ackermann, and T. J. Vogl Multi-Detector Row CT Coronary Angiography: Influence of Reconstruction Technique and Heart Rate on Image Quality Radiology, January 1, 2006; 238(1): 75 - 86. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Y. Loisance, K. Nakashima, and M. Kirsch Computer-assisted coronary surgery: lessons from an initial experience Interactive CardioVascular and Thoracic Surgery, October 1, 2005; 4(5): 398 - 401. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Falk, F. Mourgues, L. Adhami, S. Jacobs, H. Thiele, S. Nitzsche, F. W. Mohr, and E. Coste-Maniere Cardio Navigation: Planning, Simulation, and Augmented Reality in Robotic Assisted Endoscopic Bypass Grafting Ann. Thorac. Surg., June 1, 2005; 79(6): 2040 - 2047. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-H. Song, T. Ito, T. Watanabe, and H. Nakamura Multidetector Computed Tomography Versus Coronary Angiogram in Evaluation of Coronary Artery Bypass Grafts Ann. Thorac. Surg., February 1, 2005; 79(2): 585 - 588. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
