Evaluation of teleradiology for interpretation of intravenous urograms

Digital Imaging and PACS: An Update

The instant and simultaneous availability of radiologic examinations has long been of interest to referring clinicians, and certainly to radiologists, within and outside of the hospital. In daily clinical routine, the term “film management problem” has been used to describe a lack of access to this essential part of clinical information. Comprehensive picture archiving and communication systems (PACSs) have been proposed and implemented only within few institutions. This article discusses some of the most important standards, and technical and practical aspects of digital imaging and PACS.

Secure transmission of urologic images and records over the Internet

PURPOSE

Telemedicine has become a common method for the transmission of images and patient data across long distances. Our goal was to assess the efficiency and accuracy of Photomailer MD software, a store-and-forward telemedicine system, in the urologic setting.

METHODS

Photomailer MD software was loaded on two computers in the host institution, one with a T1 connection to the Internet and the other with a dial-up modem connection (24,000 bits/second), and computers at three remote sites. A total of 14 clinical cases, comprised of digitized histories and radiographic images, were sent to the remote institutions four separate times using the four transmission modes available: nonencrypted, 56-bit encryption, 128-bit encryption, and 128-bit encryption with password. The following data points were recorded: file size before and after encryption, file transmission times, and diagnostic accuracy of the remote urologists. One-way ANOVA was used to compare mean values statistically, while the z-test was used to compare diagnostic accuracies.

RESULTS

Encryption increased the file size by a mean of 37.8%, with the three encryption modes increasing file sizes by the same number of kilobytes. When a dial-up modem was used, encrypted files required a significantly longer transmission time (P < 0.05) than the unencrypted files. The same trend was seen with the T1 connection, although the differences often were not significant. When T1 transmission times were compared with modem times with other variables held constant, modem times were significantly longer (P < 0.05). Diagnostic accuracies for each of the three remote centers ranged from 85.7% to 100%. Differences in accuracy rates between attending physicians and residents were not significant.

CONCLUSIONS

Photomailer MD provides a secure, convenient, and affordable method of transmitting patient images and records via the Internet. Transmission speed was significantly greater when using a T1 line and also tended to be faster when files were not encrypted. There was no significant difference in transmission time among the three encryption modes; therefore, 128-bit encryption with a password should be used to maximize security. Diagnostic accuracies were comparable to those in the literature. In general, 640 x 480-pixel resolution was adequate for urologic diagnoses, although higher-resolution images may improve accuracy.

Teleradiology

According to the American College of Radiology, teleradiology is an electronic transmission of radiological images from one location to another for the purposes of interpretation or consultation. This article provides a historical perspective and discusses both solved and unsolved problems concerning the different elements of a teleradiology system (image acquisition, image transmission, image display, image compression). It concludes that teleradiology will in the future be a part of picture archiving and communication systems (PACS). Technical problems have for the most part been solved, with the exception of quality of gray-scale monitors. The characteristics of such a system depends on its goals.

Teleradiology: another revolution in radiology?

Teleradiology systems are rapidly being deployed by an increasing number of radiological services. Many articles have already been published on the technological developments of teleradiology but little attention has been given to its expected impact on the delivery of health care. This review article will therefore outline the historical and current technological developments of teleradiology and its potential future implementation into mainstream radiology.

Teleradiology in urology: comparison of digital image quality with original radiographic films to detect urinary calculi

PURPOSE

Teleradiology systems are now being evaluated as a mechanism to provide rapid, accurate and cost-effective diagnostic radiographs to off-site physicians. Little data are available on the role and safety of teleradiology in urology. To address these issues a personal computer based system was developed to assess the diagnostic accuracy and ease of use of transmitted digital images when evaluating for urinary calculi.

MATERIALS AND METHODS

A total of 100 plain abdominal scout films from excretory urograms performed during acute urological referrals was digitized on a laser scanner. The 10 megabyte files were transferred over public telephone lines and written to compact disks. The images were viewed on a 1280 x 1640 resolution monitor using "Imager-3D" software run on a 133 MHz. pentium personal computer with 32 megabytes of random access memory. Two faculty urologists and 2 urology fellows each looked at 50 original radiographs and 50 digital images. Diagnostic interpretations of the presence and location of calculi were recorded, and confidence in the diagnosis, assessment of image quality and diagnostic difficulty were scored using a numerical scale.

RESULTS

The accuracy for all readers was 86.5% for plain radiographs and 81.5% for digital images (p >0.2). There was no statistical difference between faculty and fellows. Diagnostic accuracy did not differ between plain films and screen images when the results were assessed with respect to image quality, diagnostic difficulty or the reader confidence in the diagnosis (p >0.1). Compared to plain films, more screen images were classified as lower image quality (60 versus 40%) and the diagnostic confidence was lower (low and medium grade 50 versus 35%), although this did not interfere with diagnostic accuracy.

CONCLUSIONS

These data imply that a high quality affordable teleradiology system is effective and accurate compared to plain films for assessing urinary calculi.

Digital radiographic imaging transfer: comparison with plain radiographs

Advances in digital imaging and computer display technology have allowed development of clinical teleradiographic systems. There are limited data assessing the effectiveness of such systems when applied to urologic pathology. In an effort to appraise the effectiveness of teleradiology in identifying renal calculi, the accuracy of findings on transmitted radiographic images were compared with those made when viewing the actual plain film. Plain films (KUB) were obtained from 26 patients who presented to the radiology department to rule out urinary calculous disease. The films were digitalized by a radiograph scanner into ARCNEMA-2 file format, compressed by a NASA algorithm, and transferred via a 28.8-kbps modern over standard telephone lines to a remote section 25 miles away, where they were decompressed and viewed on a 1600 x 1200-pixel monitor. Two attending urologists and two endourologic fellows were randomized to read either the transmitted image or the original radiograph with minimal clinical history provided. Of the 26 plain radiographic films, 24 were correctly interpreted by the fellows and 25 by the attending physicians (92% and 96% accuracy, respectively) for a total accuracy of 94% with no statistical difference (p = 0.16). After compression, all but one of the digital images were transferred successfully. The attending physicians correctly interpreted 24 of the 25 digital images (96%), whereas the fellows were correct on 21 interpretations (84%), resulting in a total 90% accuracy with a significant difference between the groups (p < or = 0.04). Overall, no statistical difference between the interpretations of the plain film and the digital image was revealed (p = 0.21). Using available technology, KUB images can be transmitted to a remote site, and the location of a stone can be determined correctly. Higher accuracy is demonstrated by experienced surgeons.

A test pattern for quality control of laser scanner and charge-coupled device film digitizers

Although clinical images provide the ultimate test of diagnostic performance for a film digitizer, such images are not generally suitable for daily quality control (QC) purposes. However, a well-designed test pattern will provide a rapid, comprehensive, objective and reproducible assessment of image quality. This pattern should evaluate various parameters of image quality, including high contrast resolution, low contrast discrimination, linearity of gray scale, geometric distortion, and noise. Furthermore, the pattern should detect light leak and film slippage, two problems commonly associated with film digitizers. The test pattern described in this manuscript was designed to provide quantitative measures of performance for a film digitizer. As part of a regular QC routine for a laser scanner or charge-coupled device digitizer, this pattern provides a simple method to identify and quantify changes in digital image quality.