R. Brennecke

Impact of routine angiographic follow-up after angioplasty

BACKGROUND There is an ongoing controversy as to whether repeat coronary angiography should be routinely performed after successful percutaneous transluminal coronary angioplasty (PTCA). METHODS We examined the 10-year outcome in 400 patients who had or had not undergone an angiographic control 6 months after successful PTCA and a subsequent event-free 6-month period. Our comparison was based on data gathered by questionnaire and telephone interview in 315 patients with (group A) and 85 patients without (group B) a routine 6-month angiographic control. Multivariate analysis (Cox model) was performed to identify predictors of adverse events. RESULTS During the 10-year follow-up period, 22 (7%) of the 315 patients in group A died, compared with 16 (19%) patients in group B (P= .003). In groups A and B, respectively, acute myocardial infarction occurred in 28 (9%) and 10 (12%) patients (not significant [NS]); coronary artery bypass grafting (CABG) was performed in 42 (13%) and 14 (16%) patients (NS); repeat PTCA was performed in 89 (28%) and 11 (13%) patients (P= .012); and serious adverse events (death, myocardial infarction, CABG) occurred in 76 (24%) and 32 (38%) patients (P= .02). Absence of a 6-month angiographic follow-up was identified as an independent predictor of death associated with a 2.7 times higher mortality rate during the 10-year follow-up period. Previous myocardial infarction increased the risk of death 2.5 times. Any increase of residual diameter stenosis by 10% was combined with a 1.4 times higher mortality rate. The chance of bypass surgery was higher in patients with multivessel disease (2.9 times), in patients with unstable angina (2.1 times), and in case of an increase of residual diameter stenosis by 10% (1.3 times). No predictor for the risk of myocardial infarction was found. Angiographic follow-up increased the likelihood of PTCA 2.5 times. CONCLUSIONS A routinely performed angiographic control 6 months after successful PTCA is associated with a significantly higher rate of repeat PTCA but, most important, is correlated with a significantly lower mortality rate during the 10-year follow-up period.

High-resolution two-dimensional echocardiography improves the quantification of left ventricular function.

It is not clear if newer echo technology prevents underestimation of ejection fraction in comparison with cineventriculography (CVG). Before undergoing CVG, 48 patients were studied with state-of-the-art, high-resolution two-dimensional echocardiography with a 3.5 MHz transducer, 96 crystals, and digital storing (slow motion and cine-loop). These data are compared with results published by us in 1983 (37 patients). Volumes were calculated using the biplane method of discs in a single area length. Exclusion criteria in 1990 were poor two-dimensional echocardiography quality (n = 6) and extra-systole during CVG (n = 4). The difference between the mean values of the angiography and echocardiography data in the new study (1990) were 4.7 ml/m2/body surface area (BSA) for the end-diastolic volume, 0.1 ml/m2 for the end-systolic volume, and 1.5% for the ejection fraction (all NS). The mean difference between both methods in the old study (1983) were -39.1 ml/m2/BSA (p less than 0.001) for end-diastolic volume, -17.4 ml/m2 (p less than 0.001) for end-systolic volume, and -7.2% for ejection fraction (p less than 0.001). The regression equation for determination of ejection fraction was y = 0.606x + 13, r = 0.803 for the old study, and y = 0.935x + 2.5; r = 0.935 for the new study. These results indicate that state-of-the-art, high-resolution two-dimensional echocardiography prevents the systematic underestimation of left ventricular volume and ejection fraction, most likely because of better delineation of endocardium in end-systole.

Comparison of QCA systems

Aims: Excellent agreement between different ‘second generation’ systems for quantitative coronary arteriography (QCA) has been found in in vitro measurements. To verify the quality and stability of QCA when used in clinical practice, three QCA systems (AWOS, Cardio, and CMS) were used in a representative set of coronary artery lesions. Methods and results: This set consisted of angiographic stenosis images of 57 patients which varied in stenosis severity and morphology. The process of image acquisition, calibration, and measurement was strictly standardized to eliminate procedural sources of error. Three observers performed QCA five times in each lesion with each QCA system. Interobserver variability was low (Dnorm 0.01–0.05 mm, Dmin 0.01–0.02 mm, %stenosis 0.3–0.7%). Values of system precision were excellent (Dnorm 0.11–0.13 mm, Dmin 0.04–0.06 mm, %stenosis 2.1–2.6%). Comparison of measurements between the three QCA systems revealed good agreement (range of mean differences for Dnorm 0.03–0.12 mm, Dmin 0.04–0.11 mm, and%stenosis 0.5–3.6%) and high correlation (corr 0.902–0.977). There was a tendency to measure smaller values for Dmin and consequently to identify more severe stenoses with the AWOS system than with the Cardio and CMS systems. All QCA results were compared to measurements done with the Brown Dodge method to reveal systematic failure of the QCA measurements. These results showed excellent agreement without any systematic deviation (mean differences for Dnorm 0.01–0.08 mm, Dmin 0.02–0.06 mm, and%stenosis 1.3–1.8%). None of the differences were statistically significant. Conclusion: We therefore conclude that using the defined version of the AWOS, Cardio, and CMS systems, there is no difference in precision or accuracy when used for QCA of coronary artery lesions.

Intravascular Ultrasound Imaging of Arterial Wall Architecture

Intravascular ultrasound (IVUS) is a promising new technique for assessing vascular morphology and structure. Controversy exists whether the three‐layer appearance of the arterial wall correctly reflects the histologic structures of the intima, media, and adventitia. We performed an experimental study to clarify the three‐layer appearance. The vessel wall architecture was analyzed by IVUS on eight different kinds of plastic cylinders, 24 normal blood vessels from pigs, and 59 human arterial segments. A distinct three‐layer appearance was observed on all the plastic cylinders when the ultrasound beam was perpendicular to the wall. A three‐layer appearance was also seen in the arterial wall, in the muscular (iliac, femoral artery) and elastic types (aorta), when the echo beam was perpendicular to the vessel wall. The three‐layer pattern was even observed on the arterial wall when the intima was removed. Furthermore, the removed intima itself provided a three‐layer image. Histologic examination showed that there was no correspondence between the IVUS three‐layer appearance and the intima, media, and adventitia. Moreover, we also performed IVUS on nine patients who suffered from aortic dissection. Intravascular ultrasonic visualization of the dissected intima showed a distinct three‐layer pattern. The pattern disappeared when: (1) the echo beam was not perpendicular to the vessel wall; (2) there was connective tissue around the vessel wall; (3) there was arterial wall calcification; or (4) the vessel wall was too thick or the distance between the ultrasound transducer and the vessel wall was too large. We conclude that the three‐layer pattern in IVUS images seems to not represent the histologic intima, media, and adventitia, but results from the reflection of echo beam at the border zones.

Comparison of image compression viability for lossy and lossless JPEG and Wavelet data reduction in coronary angiography.

Background: Lossless or lossy compression of coronary angiogram data can reduce the enormous amounts of data generated by coronary angiographic imaging. The recent International Study of Angiographic Data Compression (ISAC) assessed the clinical viability of lossy Joint Photographic Expert Group (JPEG) compression but was unable to resolve two related questions: (A) the performance of lossless modes of compression in coronary angiography and (B) the performance of newer lossy wavelet algorithms. This present study seeks to supply some of this information. Methods: The performance of several lossless image compression methods was measured in the same set of images as used in the ISAC study. For the assessment of the relative image quality of lossy JPEG and wavelet compression, the observers ranked the perceived image quality of computer-generated coronary angiograms compressed with wavelet compression relative to the same images with JPEG compression. This ranking allowed the matching of compression ratios for wavelet compression with the clinically viable compression ratios for the JPEG method as obtained in the ISAC study. Results: The best lossless compression scheme (LOCO-I) offered a mean compression ratio (CR) of 3.80:1. The quality of images compressed with the lossy wavelet-based method at CR = 10:1 and 20:1 was comparable to JPEG compression at CR = 6:1 and 10:1, respectively. Conclusion: The study has shown that lossless compression can exceed the CR of 2:1 usually quoted. For lossy compression, the range of clinically viable compression ratios can probably be extended by 50 to 100% when applying wavelet compression algorithms as compared to JPEG compression. These results can motivate a larger clinical study.

Ten-year outcome after coronary angioplasty in patients with single-vessel coronary artery disease and comparison with the results of the Coronary Artery Surgery Study (CASS).

The 10-year results of randomized trials comparing percutaneous transluminal coronary angioplasty (PTCA) in patients with single-vessel coronary artery disease (CAD) with coronary artery bypass grafting (CABG) and medical treatment are not available yet. The aim of this evaluation was to compare our 10-year follow-up results after PTCA in patients with single-vessel CAD with the 10-year follow-up results after CABG and medical treatment in the Coronary Artery Surgery Study (CASS) trial. We evaluated the clinical outcome of 509 patients with single-vessel CAD 10 years after coronary angioplasty. The data were compared with the results of 214 patients with single-vessel CAD after CABG or medical treatment from the CASS trial. End points were defined as death and myocardial infarction. Statistical evaluation was performed by life-table analysis and 2-sided Fishers exact test. The rate of survival was 86% 10 years after PTCA compared with 85% after CABG and 82% after medical treatment in patients from the CASS trial (p = NS). Survival free from myocardial infarction was 77% after coronary angioplasty, 70% after CABG, and 72% after medical treatment (p = NS). Thus, in patients with single-vessel CAD, infarct-free survival 10 years after coronary angioplasty compared favorably with the results after bypass surgery or medical treatment from the CASS trial.

Importance of calibration for diameter and area determination by intravascular ultrasound.

Background: Intravascular ultrasound (IVUS) permits quantitative assessment of the lumen diameter and area of coronary arteries. The experimental study was performed to evaluate the accuracy of diameter and area measurements.Methods and results: Lumen quantitation (lumen diameter D and cross-sectional area A) in lucite tubes (lumen diameter 2.5 to 5.7 mm, Plexiglasℳ) was performed using a mechanical IVUS system (HP console, 3.5F catheter, Boston Scientific, 30 MHz). The influence of fluid type (blood, water and saline solution), fluid temperature (20°C/37°C), catheter to catheter variation, gain setting and ultrasound frequency (12, 20 and 30 MHz) was determined. In blood at 20°C there was a constant deviation of the measured diameter from the true luminal diameter of −0.29 ± −0.04 mm (p<0.06). In water and saline solution at 20°C the mean deviation from true diameter was −0.21 ± −0.06 mm (p<0.06). At 37°C, the deviation in blood was greater than at 20° (−0.34 ± −0.02 mm) which is >10% in a 3mm tube (p<0.06). Three of the ten catheters tested in water at 20°C underestimated true diameter by more than −0.3 mm. The deviation from true diameter (5mm tube) with varying gain settings was −0.14 mm to −0.23 mm compared to −0.19 mm at standard settings (p>0.288). At 12 MHz diameter measured was over-estimated. The error in absolute area estimation increased with increasing diameter tested in blood at 37°C (−1.21 to −2,72mm2), whereas the relative error ([Measured Area-True Area]/True Area × 100 [%]) was more striking at smaller diameters (up to −25% in the 2.5 mm tube).Conclusion: Luminal diameters and areas are underestimated by this particular IVUS system. When IVUS imaging and measurements are made during coronary interventions this error should be taken into account with regard to appropriate sizing of the device and the assessment of the postprocedure result. Because systematic errors might also occur in other IVUS systems (not tested in this study), it is advisable to ensure that each system is validated prior to clinical use, especially when exact measurements are required.

A Multiprocessor-System for the Acquisition and Analysis of Video Image Sequences

We give a preliminary report on the organization, architecture and programming of an image processing system optimized for the high speed storage, handling and analysis of digitized time-varying video image sequences. Typical known video processor architectures are characterized by dedicated data paths. Spatial multiplexing is used for system reconfiguration. The multiprocessor architecture described here uses a time-multiplexed synchronous bus for easy system expansion. Control and computational tasks are distributed among several levels of software and corresponding hardware to simplify efficient programming. The basic data structure supported is the data stream. Data streams are assembled and distributed under programm control. Performing I/0- and computational tasks can be visualized as passing these structured data streams through a chain of programmed processors. We expect that this concept, which is supported by hardware, will simplify the efficient programming of concurrent I/0- and computational processes required for fast execution of image sequence processing tasks. Applications from medical X-ray image processing are given.

A framework for PACS development in cardiology

The authors develop a framework for the integration of cardiology into hospitalwide picture archiving and communication systems (PACSs) by the definition of user requirements in cardiology image information systems. These user requirements have to be based on the consensus of users and industry. A topdown layered scheme is proposed for the development of documents based on user requirements. The structure of committees within the European Society of Cardiology that are working in this field is discussed.<<ETX>>

Improving the applicability of myocardial densitometry and parametric imaging by extended automated densogram analysis

In clinical applications the analysis of X-ray contrast densograms acquired in regions of interest (ROIs) over the myocardium is disturbed by many complex factors. For this reason we acquire redundant densogram information for quality control before extracting densitometric parameters. In our approach, initially some stable measures of quality for densograms are used to lower the influence of poor quality densograms by a quality weighted averaging. For example a shape quality measure, Q1, is calculated using regions of optimal and minimal acceptable quality defined with respect to a prototype densogram. Not a few myocardial ROIs yield densograms that differ from single-source densograms (SSDs) due to e.g. superposition of different perfusion beds or the position of the ROI relative to the coronary sinus or stenoses. This might result in a densogram shape with oscillating or plateau behavior. For densograms of a such general shape many parameters defined in the usual way do not depend smoothly on the densogram values. The conventional definitions of some parameters (appearance time, rise time) are therefore extended for application to multi-maxima densograms as well as to SSDs. These new methods are evaluated using digitized clinical angiocardiograms and are applied to parametric imaging (pixeldensograms) in a slightly modified way. Taking into account the densogram quality, its shape and its origin results in a considerable improvement both for densitometry and parametric imaging of myocardial perfusion.