INTERPRETATION OF HEAD CTs ON PAC SYSTEMS: COMPARISON OF 1024x768 TO 2048x2048 MONITORS.

Maria Tzalonikou, M.D., Dimitrios Spigos, M.D., Eric Bourekas, M.D., Greg Christoforidis, M.D.

Department of Radiology
Ohio State University Hospitals,
Columbus (Ohio),  U.S.A.

Purpose:

The aim of this study was to evaluate the accuracy of 1024x786 personal computer (PC) display as compared to 2048x2048 workstation picture archiving and communication (PACS) display system in the evaluation of emergency head computed tomograms (CT) in our totally filmless Emergency Department.

Materials and Methods:

One radiologist reviewed 390 head CTs during a 3 month period. All these studies were first interpreted on the 1024x768 PC system. This PC system was was intented for the use of the clinicians in the ED. The films from this interpretation were compared to those of the official report interpreted on the 2048x2048 Radiology Department PACS workstation. This system was considered the gold standard and is used for final interpretation by radiologists.

Results:

Out of the 390 cases of head CTs, 167 were negative. Forty were essentially normal, with incidental findings such as mucoperiosteal thickening of the paranasal sinuses. The remainder 215 head CTs had positive findings.

Of the 215 positive CTs, there were 13 cases of primary or secondary brain tumors, 44 cases of infarcts, 9 cases of lacunar infarcts and 30 cases of white matter changes. Thirty-six patients presented with various kinds of bleeding: nine cases of subdural hemorrhage, 3 cases of sybarrachnoid hemmorhage, 2 of intraventricular hemorrhage, 12 of brain contusions and 14 cases of hematomas. Nine patients presented either hydrocephalus or mild ventricular dilatation. Fifty two patients presented brain atrophy, while 17 patients had postoperative changes. Eleven patients had vascular lesions, among which 5 presented brain aneurysms. Six had questionable findings. A total of 105 cases presented with other findings, such as osseous abbnormalities, etc. Out of these 215 positive CTs, the pathologic findings were not evident in 3 cases (1.39%) on the 1024x 768 PC that were visible on the PAC system. All of them were brain contusions, 2 of which were located in the frontal lobe behind the frontal sinuses, while the third one was more posterior in the left frontal lobe.

Conclusion:

In general, the accuracy of 1024x768 PC is acceptable for most pathologic entities. The PC system may fail for any subtle abnormality that is relatively isodense. For the ED clinicians this review system is acceptable. They have immediate access to the images directly for their own review. For a final official Radiologist interpretation, the PC system may not prove adequate for subtle pathology and the standard PAC system is essential.

The corresponding author is

Dr. D. Spigos,
Professor and Chairman of Radiology,
Department of Radiology,
S-209 Rhodes Hall, Ohio State University Hospitals,
410 West 10th Ave.,
Columbus Ohio, 43210 U. S. A.
TEL: 001-614-2934589
FAX: 001-614-2938221

Oral presentation at EuroPACS'98, Barcelona, Spain