Effects of MPEG compression on the quality and diagnostic accuracy of digital echocardiography studies

Pitfalls in diagnosing PFO: characteristics of false-negative contrast injections during transesophageal echocardiography in patients with patent foramen ovales

BACKGROUND

Contrast injections during transesophageal echocardiography for patent foramen ovale (PFO) detection may be false negative. The characteristics of false-negative injections were studied retrospectively.

METHODS

Contrast transesophageal echocardiography was analyzed for the presence or absence of two characteristics: leftward bulging of the interatrial septum and dense contrast filling of the region of the right atrium adjacent to the interatrial septum.

RESULTS

Two hundred forty-seven injections were administered to 14 patients with PFOs, and 130 (53%) were false negative. The absence of either characteristic during a single injection resulted in low sensitivity for PFO detection of 7%. When the two characteristics were present at the same time, the sensitivity for PFO detection after a single injection was as high as 95%.

CONCLUSIONS

The simultaneous presence of both leftward bulging of the interatrial septum and dense contrast filling of the region in the right atrium, adjacent to the interatrial septum, is a prerequisite for PFO detection, and if either characteristic is missing, the injection is inconclusive.

Clinically Compressed Digital Echocardiography: A Patient-safe Alternative to Videotape Review

Introduction: Digital storage of echocardiographic data offers logistical advantages over videotape archival. However, limited information is available on the accuracy of clinically compressed digitised examinations, an important consideration for patient safety. Materials and Methods: Transthoracic echocardiograms of 520 consecutive patients were prospectively acquired digitally and on videotape. Two echocardiologists, in consensus, reported studies in both formats sequentially. Using the videotape as a reference, the significance of any reported differences was graded from both imaging and clinical standpoints, and the reasons for these differences identified. Results: From an imaging perspective, differences between digital and videotaped studies were absent or minor in 459 cases (88%), fairly significant in 55 (11%) and very significant in 6 (1%). The main reasons for the observed differences were inadequate acquisition of optimal views (59%), an insufficient number of acquired cardiac cycles (25%) and suboptimal image quality (9%). These differences were considered to be of possible or definite clinical importance in 21 (4%) and 8 (2%) cases, respectively. In multinominal logistic regression models, the only independent predictor of significant difference between digitised and videotaped images was study complexity. Regardless of case complexity, most diagnostic errors arising from digital review were attributable to technical failure rather than observer error. Conclusions: The potential for important errors arising from exclusive reporting of clinically compressed digital echocardiograms is small. Digital echocardiography, as practiced in a routine clinical setting, offers a patient-safe alternative to videotape review. Key words: Clinical implications, Digital echocardiography, Patient safety, Videotape

A Protocol for the Assessment of Diagnostic Accuracy in Tele-echocardiography Imaging

Tele-echocardiography could be a useful means for investigating heart pathologies on remotely-located patients. At present, the main drawback with tele-echocardiography is the lack of a thorough protocol for the assessment of the diagnostic accuracy of the transmitted images. Diagnostic accuracy in tele-echocardiography is not only a function of quantitative parameters but also of the subjective decision of the operator depending on his/her a priori knowledge based on complex internal models. In the framework of three Italian projects, we defined and validated a wide-ranging protocol that considers not only the most common quantitative parameters in medical imaging, but also the use of a dynamic phantom, and subjective/partially subjective evaluations. The validation of tele-echocardiography systems chosen from those evaluated in the projects showed that the protocol was feasible. It permitted access to the degradation of the transmitted images and correlate quantitative and subjective analyses. The protocol was found suitable for tele-echocardiography systems but also for other tele-imaging applications, where medical decision making is based on dynamic images.

High-grade video compression of echocardiographic studies: a multicenter validation study of selected motion pictures expert groups (MPEG)-4 algorithms

BACKGROUND

Large files produced by standard compression algorithms slow down spread of digital and tele-echocardiography. We validated echocardiographic video high-grade compression with the new Motion Pictures Expert Groups (MPEG)-4 algorithms with a multicenter study.

METHODS

Seven expert cardiologists blindly scored (5-point scale) 165 uncompressed and compressed 2-dimensional and color Doppler video clips, based on combined diagnostic content and image quality (uncompressed files as references). One digital video and 3 MPEG-4 algorithms (WM9, MV2, and DivX) were used, the latter at 3 compression levels (0%, 35%, and 60%).

RESULTS

Compressed file sizes decreased from 12 to 83 MB to 0.03 to 2.3 MB (1:1051-1:26 reduction ratios). Mean SD of differences was 0.81 for intraobserver variability (uncompressed and digital video files). Compared with uncompressed files, only the DivX mean score at 35% (P = .04) and 60% (P = .001) compression was significantly reduced. At subcategory analysis, these differences were still significant for gray-scale and fundamental imaging but not for color or second harmonic tissue imaging. Original image quality, session sequence, compression grade, and bitrate were all independent determinants of mean score.

CONCLUSIONS

Our study supports use of MPEG-4 algorithms to greatly reduce echocardiographic file sizes, thus facilitating archiving and transmission. Quality evaluation studies should account for the many independent variables that affect image quality grading.

A low-cost digital filing system for echocardiography data with MPEG4 compression and its application to remote diagnosis

The high cost of digital echocardiographs and the large size of data files hinder the adoption of remote diagnosis of digitized echocardiography data. We have developed a low-cost digital filing system for echocardiography data. In this system, data from a conventional analog echocardiograph are captured using a personal computer (PC) equipped with an analog-to-digital converter board. Motion picture data are promptly compressed using a moving pictures expert group (MPEG) 4 codec. The digitized data with preliminary reports obtained in a rural hospital are then sent to cardiologists at distant urban general hospitals via the internet. The cardiologists can evaluate the data using widely available movie-viewing software (Windows Media Player). The diagnostic accuracy of this double-check system was confirmed by comparison with ordinary super-VHS videotapes. We have demonstrated that digitization of echocardiography data from a conventional analog echocardiograph and MPEG 4 compression can be performed using an ordinary PC-based system, and that this system enables highly efficient digital storage and remote diagnosis at low cost.

Forward and store telemedicine using Motion Pictures Expert Group: a novel approach to pediatric tele-echocardiography

BACKGROUND

Live transmission of echocardiograms over integrated services digital network lines is accurate and has led to improvements in the delivery of pediatric cardiology care. Permanent archiving of the live studies has not previously been reported. Specific obstacles to permanent storage of telemedicine files have included the ability to produce accurate images without a significant increase in storage requirements.

OBJECTIVE

We evaluated the accuracy of Motion Pictures Expert Group (MPEG) digitization of incoming video streams and assessed the storage requirements of these files for infants in a real-time pediatric tele-echocardiography program.

RESULTS

All major cardiac diagnoses were correctly diagnosed by review of MPEG images. MPEG file size ranged from 11.1 to 182 MB (56.5 +/- 29.9 MB).

CONCLUSIONS

MPEG digitization during live neonatal telemedicine is accurate and provides an efficient method for storage. This modality has acceptable storage requirements; file sizes are comparable to other digital modalities.

A Perioperative Echocardiographic Reporting and Recording System

Advances in video capture, compression, and streaming technology, coupled with improvements in central processing unit design and the inclusion of a database engine in the Windows operating system, have simplified the task of implementing a digital echocardiographic recording system. I describe an application that uses these technologies and runs on a notebook computer.

Digital video for the documentation of colonoscopy

BACKGROUND

The aim of this study was to determine whether digital video is suitable for the documentation of colonoscopy. Standards are required for the visual documentation of endoscopic findings and to optimize image quality while limiting file size and bandwidth requirements.

METHODS

Video recordings of colonoscopy procedures were encoded using a common video compression method at selected data rates and resolutions. Twelve reviewers were selected, each of whom was assigned 8 video review sessions, each consisting of 5 colonoscopy procedures. The reviewers rated the following: level of confidence that the cecum was demonstrated, subjective quality of the video compared with actual videocolonoscopy, and whether the video was of "diagnostic quality."

RESULTS

Reviewers were confident that the cecum was demonstrated in all cases except at the lowest data rate. The 1.0 Mbps standard interchange format video provided an optimal balance between quality and file size.

CONCLUSIONS

For the documentation of colonoscopy, 1.0 Mbps is acceptable and results in a file size of 7.5 Mbytes/min, which is manageable for most modern hospital and telehealth networks.

Web-based delivery of medical multimedia contents using an MPEG-4 system

Moving picture expert group compression standard version 4 (MPEG-4) is a standard for video coding aimed at multimedia applications. MPEG-4 was developed to enable high compression rate in a low bitrate transmission via the Internet or mobile telecommunications. Although these characteristics of MPEG-4 are suitable for telemedicine, little is known about the possibility of using this technology in the field of telemedicine. We evaluated the quality of MPEG-4-encoded medical video streams and compared them with original analogue videos and audio-video-interleave (AVI) files. Although MPEG-4 video streams have the advantage of small file size, they were found to be inferior to original videos and AVI files in terms of smoothness of motion pictures, sharpness of images and clearness of sound. Illegibility of characters was a major problem in MPEG-4 files. The score for total impression of MPEG-4 files was significantly lower than those for AVI files. The results of this study suggest that the quality of MPEG-4-encoded video streams is not adequate for telemedicine.

MPEG digital compression and analogue videotape: a comparison of moving images and electroencephalogram data in epileptic patients

PRIMARY OBJECTIVES

The objective of this study was to compare the clinical usefulness, in the field of epileptology, of digital moving images and electroencephalogram (EEG) waveforms by Motion Pictures Expert Group compression algorithms (MPEG-1 and MPEG-2) to that of conventional analogue recording.

RESEARCH DESIGN AND METHODS

Three epileptic seizure scenes consisting of moving images and the corresponding EEG waveforms in an epileptic patient were selected as the images to be evaluated. Each scene was recorded using MPEG-1, MPEG-2 and videotape. Ten doctors used six criteria to evaluate the quality of moving images, EEG data and audio.

MAIN OUTCOMES

Analysis of variance and Bonferroni tests indicated that the image quality of MPEG-2 was superior to that of MPEG-1 or videotape for all criteria. Furthermore, MPEG-2 obtained much higher scores in EEG waveform quality than did the other modalities.

CONCLUSIONS

Our findings suggested that data from MPEG-2 images will lead to more precise diagnosis and treatment decision-making than data from analogue videotape recordings.

Comparison of diagnostic quality of motion picture experts group-2 digital video with super VHS videotape for echocardiographic imaging

BACKGROUND

Motion Picture Experts Group-2 (MPEG2) is a broadcast industry standard that allows high-level compression of echocardiographic data. Validation of MPEG2 digital images compared with super VHS videotape has not been previously reported.

METHODS

Simultaneous super VHS videotape and MPEG2 digital images were acquired. In all, 4 experienced echocardiographers completed detailed reporting forms evaluating chamber size, ventricular function, regional wall-motion abnormalities, and measures of valvular regurgitation and stenosis in a blinded fashion. Comparisons between the 2 interpretations were then performed and intraobserver concordance was calculated for the various categories.

RESULTS

A total of 80 paired comparisons were made. The overall concordance rate was 93.6% with most of the discrepancies being minor (4.1%). Concordance was 92.4% for left ventricle, 93.2% for right ventricle, 95.2% for regional wall-motion abnormalities, and 97.8% for valve stenosis. The mean grade of valvular regurgitation was similar for the 2 techniques.

CONCLUSIONS

MPEG2 digital imaging offers excellent concordance compared with super VHS videotape.

Digital echocardiography 2002: now is the time

The ability to acquire echocardiographic images digitally, store and transfer these data using the DICOM standard, and routinely analyze examinations exists today and allows the implementation of a digital echocardiography laboratory. The purpose of this review article is to outline the critical components of a digital echocardiography laboratory, discuss general strategies for implementation, and put forth some of the pitfalls that we have encountered in our own implementation. The major components of the digital laboratory include (1) digital echocardiography machines with network output, (2) a switched high-speed network, (3) a high throughput server with abundant local storage, (4) a reliable low-cost archive, (5) software to manage information, and (6) support mechanisms for software and hardware. Implementation strategies can vary from a complete vendor solution providing all components (hardware, software, support), to a strategy similar to our own where standard computer and networking hardware are used with specialized software for management of image and measurement information.

Evaluation of valvular regurgitation severity using digital acquisition of echocardiographic images

BACKGROUND

Digital acquisition is a technique for storing echocardiographic data that offers advantages over conventional videotape (VT); however, limited information is available on its accuracy for the evaluation of valvular regurgitation.

METHODS

We evaluated 102 patients with at least 1 regurgitant lesion. Data were obtained on VT and in 1 cardiac cycle stored digitally (1C). To assess for incremental improvement with acquisition of multiple cycles, digital images were also acquired with 2 (2C) or 3 cardiac cycles (3C). Both digital and VT images were graded for regurgitant severity as absent, trivial, mild, moderate, or severe. Kappa statistics were used to assess agreement.

RESULTS

A total of 171 valvular regurgitant lesions (mild or greater) were evaluated. The overall agreement between 1C and VT images was kappa = 0.61. With multiple cycle acquisition, there was no improvement in agreement (kappa = 0.56 and 0.57 for 2C and 3C, respectively). When subgrouped, the level of agreement between 1C and VT was slightly lower for the aortic valve than for the mitral or tricuspid valves (kappa = 0.49, 0.63, 0.64, respectively).

CONCLUSION

The 1C technique has substantial agreement and correlation with standard VT for the evaluation of regurgitant lesions with the use of color flow Doppler. The acquisition of multiple cardiac cycles does not provide incremental improvement over single beat acquisition.

Telecardiology: potential impact on acute care

Echocardiography is often used to diagnose and exclude important cardiac diagnoses in adults and children. Evolving telemedicine technology has the potential to improve access to echocardiography diagnoses in the intensive care unit, emergency room, and newborn nursery. The two primary modes of telemedicine practice are "store and forward" and "real-time" videoconferencing. A digital echocardiogram (often several one cardiac cycle loops) can be stored at one site and forwarded across a telemedicine network to a receiving station for review at a later time. Pediatric cardiologists often favor "real-time" telemedicine because of the ability to guide sonographers with limited experience in congenital heart disease. A complete telemedicine system requires a modified computer, a low- or high-speed connection, and telemedicine inputs. Several adult and pediatric clinical studies have shown telemedicine to be accurate and cost-effective, improve patient care, enhance echocardiogram quality and sonographer proficiency, and promote practice expansion. Obstacles to widespread implementation of telemedicine include lack of standardization of telemedicine components, confusing legal issues and licensure requirements, and poor reimbursement.

Accuracy and cost- and time-effectiveness of digital clip versus videotape interpretation of echocardiograms in patients with valvular disease

BACKGROUND

Although digital and videotaped images are known to be comparable for the evaluation of left ventricular function, their relative accuracy for assessment of more complex anatomy is unclear. We sought to compare reading time, storage costs, and concordance of video and digital interpretations across multiple observers and sites.

METHODS

One hundred one patients with valvular (90 mitral, 48 aortic, 80 tricuspid) disease were selected prospectively, and studies were stored according to video and standardized digital protocols. The same reviewer interpreted video and digital images independently and at different times with the use of a standard report form to evaluate 40 items (e.g., severity of stenosis or regurgitation, leaflet thickening, and calcification) as normal or mildly, moderately, or severely abnormal. Concordance between modalities was expressed at kappa. Major discordance (difference of >1 level of severity) was ascribed to the modality that gave the lesser severity. CD-ROM was used to store digital data (20:1 lossy compression), and super-VHS videotape was used to store video data. The reading time and storage costs for each modality were compared.

RESULTS

Measured parameters were highly concordant (ejection fraction was 52% +/- 13% by both). Major discordance was rare, and lesser values were reported with digital rather than video interpretation in the categories of aortic and mitral valve thickening (1% to 2%) and severity of mitral regurgitation (2%). Digital reading time was 6.8 +/- 2.4 minutes, 38% shorter than with video (11.0 +/- 3.0, range 8 to 22 minutes, P <.001). Compressed digital studies had an average size of 60 +/- 14 megabytes (range 26 to 96 megabytes). Storage cost for video was A$0.62 per patient (18 studies per tape, total cost A$11.20), compared with A$0.31 per patient for digital storage (8 studies per CD-ROM, total cost A$2.50).

CONCLUSION

Digital and video interpretation were highly concordant; in the few cases of major discordance, the digital scores were lower, perhaps reflecting undersampling. Use of additional views and longer clips may be indicated to minimize discordance with video in patients with complex problems. Digital interpretation offers a significant reduction in reading times and the cost of archiving.

Transition to an all-digital echocardiography laboratory: a large, multi-site private cardiology practice experience

Acquisition, interpretation, and storage of digital echocardiographic images has many advantages over the standard videotape-based method. Archival, transmission, and comparative interpretation are all optimized with digital echocardiography. A study performed at one site can be immediately available for viewing and analysis at another site by means of standard data transfer technology. Echocardiograms can be interpreted in the context of prior studies, which are readily available for side-by-side comparison. The transition to an all-digital laboratory involves the commitment of persons at multiple levels in the cardiology practice, including administrators, information technology specialists, sonographers, and physicians. Quality of patient care, use of physicians' and sonographers' time, and long-term financial benefit are all areas where improvement may be realized with the use of digital echocardiography. We present our experience in the development of an all-digital echocardiography laboratory, and we conclude that digital echo-cardiography is practical and can be implemented readily in a clinical setting. We performed several correlative analyses during this transition to validate the consistency and accuracy of digital interpretation compared with those of analog methods. The transition process from analog (videotape) to digital, including full wide area network exchange, took approximately 8 months. As technology advances, issues surrounding storage, comparison, and acquisition formats will continue to develop. We hope that our experience will help others make the transition to the digital environment and benefit from the ease of image access, the ability to comparatively interpret echocardiograms, and the superior image quality afforded by this advancement.

Electronic transmission of digital echocardiographic studies: effects of MPEG compression

The acquisition, storage and retrieval of digital echocardiographic studies greatly facilitates image review and quantitation and permits the transmission of studies electronically. However, the considerable size of digital echocardiographic data files makes transmission over existing networks slow and impractical. Reduction in the size of these data files can be accomplished with digital image compression. We sought to evaluate the effects of MPEG-1 compression on the transfer time of digital echocardiographic studies over currently available network connections. Ninety consecutive routine clinical echocardiographic studies were randomly compressed at one of three compression ratios 60:1, 80:1, or 120:1 and sent to a receiving terminal using simulated transmission rates. Compression of digital echocardiographic studies at these ratios which, have been shown to maintain diagnostic image quality, reduced the size of digital echocardiographic studies to less than 1% of their original sizes which allowed transmission of echocardiographic studies over networks using 3ISDN or T1 lines with minimal waiting time.