IT services in a completely digitized radiological department: value and benefit of an in-house departmental IT group

To analyze the benefit of a departmental IT group in comparison to support by hospital IT groups or system manufacturers in a completely digitized radiological department. The departmental IT group comprises a fulltime IT specialist, two student assistants and four clinical employees participating 1 day/week. For 18 months IT problems were quantified and specified according to urgency, responsibility and affected system by use of an intranet-based reporting system. For each IT service provider the performance and duration of problem solution was evaluated. In 18 months 3,234 IT problems emerged. 88.7% were solved by the departmental IT group. In 474 cases (14.7%) a solution within 2 h was required. The departmental IT group solved 35.8% within 30 min, system manufacturers needed 18 h 38 min in mean. The departmental IT group solved 90.2% of the problems within a time limit. System manufacturers met the limit in 60.1% with a mean duration of 7 days 21 h. In 6.7% of the cases, support by system manufacturers was indispensable. A considerable proportion of IT problems in completely digitized radiological departments can be solved by a departmental IT group, providing a fast and cost-efficient first-level IT support with effective prevention of major breaks in the workflow. In a small number of cases support by system manufacturers remains necessary.This work was orally presented on ECR 2004 in Vienna.     

Multi-detector computed tomography and 3-dimensional imaging in a multi-vendor picture archiving and communications systems (PACS) environment

RATIONALE AND OBJECTIVES: To show the impact of the introduction of multi-detector computed tomography (CT) on radiologic workflow and to demonstrate how this reflects on picture archiving and communications systems (PACS) requirements. MATERIALS AND METHODS: Production measurements were obtained from different CT scanners (first two single-slice CT scanners; from December 2001 single and 4-slice CT; from April 2002 single and 16-slice CT) in number of patients from the radiologic information system. Implications on our PACS were recorded in terms of images and studies stored. Furthermore, our PACS design was made so that optimal use of 3-dimensional imaging within the radiologic workflow was possible. Finally, the number of non-diagnosed studies were recorded every day since the start of the transition to a filmless radiology department. RESULTS: This PACS design achieved a high level of integration between simple viewing and advanced 3-dimensional imaging and is optimized for handling large amounts of data. Overall increase of patients scanned with CT from January 2002-December 2003 was 54%. The number of series increased by 286% from December 2001-April 2003 and by 130% from April 2002-December 2003. From January 2002-February 2003, the number of images per patient increased from 175 to 450 (157%). Non-diagnosed studies decreased from about 100-120 before to practically zero after PACS implementation. CONCLUSION: PACS significantly increases productivity because of availability of the images and elimination of certain manual tasks. These results show that although the amount of examinations increases significantly with the introduction of MDCT, simultaneous introduction of PACS and filmless operation allows radiologists to handle the growth in workload.     

Diagnostic performance of liquid crystal and cathode-ray-tube monitors in brain computed tomography

The aim of this study was to evaluate feasibility of reporting brain CT examinations on liquid crystal display (LCD) flat-screen monitors vs state-of-the-art cathode-ray-tube (CRT) monitors. Ninety-five brain CT examinations of 95 patients were displayed on Picture archiving and communications system (PACS) workstations equipped either with a dedicated medical imaging LCD colour monitor or on a high-resolution CRT which is used for routine reporting of CT, MRI and digital radiography images in our institution. Fifty cases were negative and 45 cases were positive for early brain infarction (EBI), the latter being defined by a combination of one or more signs: dense artery; hypodensity of brain parenchyma; and local brain swelling verified by control scans. Ten radiologists had to rate presence or absence of EBI on a five-point scale. Ratings were evaluated by CORROC2 ROC software and areas under the ROC curve (A_z) were computed. Significance of differences between the two viewing conditions were evaluated with Wilcoxon test. Mean A_z of the ten observers was 0.7901 with LCD vs 0.7695 with CRT which did not show statistical significance (p=0.2030). In the setting investigated, reporting of CT studies from high-performance LCD monitors seems feasible without significant detriment to diagnostic performance.     

Performance of Web-based image distribution: client-oriented measurements

The aim of this study was to define a clinically suitable personal computer (PC) configuration for Web-based image distribution and to assess the influence of different hard- and software configurations on the performance. Through specially developed software the time-to-display (TTD) for various PC configurations was measured. Different processor speeds, random access memory (RAM), screen resolutions, graphic adapters, network speeds, operating systems and examination types (computed radiography, CT, MRI) were evaluated, providing more than half a million measurements. Processor speed was the most relevant factor for the TTD; doubling the speed halved the TTD. Under processor speeds of 350 MHz, TTD mostly remained above 5 s for 1 CR or 16 CT images. Here Windows NT with lossy compression were superior. Processor speeds of 350 MHz and over delivered TTD <5 s. In this case Windows 2000 and lossless compression were preferable. Screen resolutions above 1280×1024 pixels increased the TTD mainly for CR images. The RAM amount, network speed and graphic adapter did not have a significant influence. The minimum threshold for clinical routine is any standard off-the-shelf PC better than Pentium II 350 MHz, 128 MB RAM; hence, high-end PC hardware is not required.     

图像存储与传输系统在放射科的建设与应用

目的:探讨在组建图像存储与传输系统(picture archiving and communication system,PACS)过程中的总体设计与软件体系结构。方法:把具有医学数字成像及通讯(digital imaging and communication in medicine,DICOM)标准接口或非DlCOM标准接口的影像设备进行联网,制定资源共享,系统存储的解决方案,建立典型的医院放射科PACS系统,连接目前医院现有的设备:磁共振、数字乳腺、KODAK及富士X线摄影,GE计算机X线摄影及GE-CT、GE-DSA,解决管理及存储问题。结果:建立了典型的医院放射科PACS系统,连接了目前医院的现有设备,实现了放射科初步的无胶片化方式,建立起了影像数据中心,使用起来较为得心应手,方便了医生。结论:图像存储与传输系统在放射科的建设与初级应用,提高了放射科室的QA(质量管理)、QC(质量控制动)的控制水平,提高了工作效率。     

北京大学第三医院医学生殖中心设计综述

结合北京大学第三医院医学生殖中心净化工程设计与施工配合的工作经历,重点分析了医学辅助生殖中心的平面设计布置、流程及细节的设计。     

Accuracy of digital imaging interpretation on an 1K × 1K PC-based workstation in the emergency department

The purpose of this study was to evaluate the accuracy of interpretation of digital radiographs on 1K × 1K PC-based workstations and compare it to that of 2K × 2K workstations in our filmless emergency department. We reviewed 1013 studies, of which 404 were chest, 133 abdominal, and 476 musculoskeletal. All studies were interpreted prospectively on the 1K × 1K workstation, and the findings were compared to those on the 2K × 2K diagnostic workstation which served as the control. Among all the studies performed, 515 were negative (255 chest, 91 abdominal, and 169 musculoskeletal). Another 498 had positive findings (149 chest, 42 abdominal, and 307 musculoskeletal). When the cases were reviewed on the 1K × 1K monitor, the diagnosis was missed in 5 of the 149 positive chest cases (3.35 %), in 5 of the 307 positive musculoskeletal cases (1.63 %), and in none of the 42 positive abdominal cases (0 %). In total, we misinterpreted 10 of the 498 positive cases (2.01 %). The miss rate of the 1K × 1K clinicians' review station is low and acceptable for its stated purpose. However, we doubt that it can be used for primary diagnostic purposes.     

PACS系统存储方案的设计原则和实现策略

结合我院ＰＡＣＳ存储方案的设计实践和我们论证过程，探讨适合国情的ＰＡＣＳ存储方案的设计原则和优化的选择。材料与方法，分析和论证具有自动装载功能的库类海量存储设备及其相关系统，如磁带库，光盘库等。涉及的存储介质为：Ｄ磁盘、ＭＯＤ（ＭａｇｎｅｔｉｃＯｐｔｉｃＤｉｓｋ）、ＣＤ／ＤＶＤ和ＤＬＴ（ＤｉｇｉｔａＬｉｎｅａｒＴａｐｅ）；通过论证数据产生总量及增长比率确定存储容量及要求，分析影像调用频率及间期定影     

A comparative study of a digital radiography system

The present study aimed to examine the efficiency of a new digital radiography system that was installed in the Royal Adelaide Hospital in September 1997, as compared to the existing conventional radiography system. A total of 55 examinations were observed over a 3-week period in January, and these consisted of 18 digital and 10 conventional chest examinations, and 27 conventional orthopaedic examinations. These were combined with 18 digital orthopaedic examinations recorded from a prior study. Total examination time was broken into several components, of which reporting time was of the most interest. The mean reporting times for digital and conventional chest examinations were 17 and 25 min, respectively, a significant (P < 0.1) 8-min difference. The orthopaedic examinations revealed mean reporting times of 8 and 26 min for digital and conventional systems, respectively; a significant (P < 0.001) 18-min difference. These results demonstrate that the digital system is a faster, more efficient system for the reporting of X-rays.