Image retrieval takes seconds even on 'slow' PCsMedical imaging has embraced the PC, in many cases abandoning proprietary computing platforms. This is especially apparent in Web-based teleradiology and PACS efforts that extend
Image retrieval takes seconds even on 'slow' PCs
Medical imaging has embraced the PC, in many cases abandoning proprietary computing platforms. This is especially apparent in Web-based teleradiology and PACS efforts that extend beyond the radiology department. But the term "PC" is sufficiently generic as to raise a question about how good a PC is good enough. The answer can be critical when configuring IT hardware.
Researchers at the Johann Wolfgang Goethe University Hospital in Frankfurt, Germany, looked into this question, finding that even relatively unsophisticated hardware can meet performance requirements for Web-based image review. They tested seven PCs spanning three generations of processor (Pentium I, II, and III), using differing configurations of RAM, network, operating system, and graphic adaptor. They measured the time-to-display (TTD) for CR, CT, and MR images from five separate examinations on each PC, while also altering screen resolution settings and switching from lossy to lossless compression.
Processor power was the major limiting factor in performance. Doubling the processor speed halved the TTD, though even PCs with relatively modest processing power (350 MHz and above) managed to retrieve either one computed radiograph or 16 CT images within five seconds. Upping screen resolution beyond 1280 x 1024 pixels slowed the TTD most significantly for CR images.
"The performance level of PC we are talking about here is not that high," said Dr. Bjorn Bergh, who trained as a radiologist in Berlin before becoming IT director at the Frankfurt hospital. "We were actually astonished that you could use (a low-powered) PC."
Bergh warned, however, that PC configuration is not the only factor affecting the performance of Web-based systems. Network speed can also influence TTD.
"Hospitals with a significantly slower network than the one we were using might find they have an information bottleneck in the network, not on the PC," he said.
MRI-Based AI Radiomics Model Offers 'Robust' Prediction of Perineural Invasion in Prostate Cancer
July 26th 2024A model that combines MRI-based deep learning radiomics and clinical factors demonstrated an 84.8 percent ROC AUC and a 92.6 percent precision-recall AUC for predicting perineural invasion in prostate cancer cases.
Breast MRI Study Examines Common Factors with False Negatives and False Positives
July 24th 2024The absence of ipsilateral breast hypervascularity is three times more likely to be associated with false-negative findings on breast MRI and non-mass enhancement lesions have a 4.5-fold likelihood of being linked to false-positive results, according to new research.
Can Polyenergetic Reconstruction Help Resolve Streak Artifacts in Photon Counting CT?
July 22nd 2024New research looking at photon-counting computed tomography (PCCT) demonstrated significantly reduced variation and tracheal air density attenuation with polyenergetic reconstruction in contrast to monoenergetic reconstruction on chest CT.