Overnight Preliminary Interpretations of CT and MR Images by Radiology Residents in ER: How Accurate Are They?

PURPOSE: At many institutes in Korea, preliminary interpretations of after-hours CT and MR images are performed by radiology residents, with the attending radiologist's reviewing the interpretations the next day. The purpose of this study was to assess the rate of discrepancy between residents' interpretations and the final interpretations performed by attending radiologists. METHODS: We reviewed the interpretations of 1381 CT and 404 MRI scans that were obtained at the emergency department of our institute over three months. Any discrepancies between the preliminary and final interpretations were categorized as either major or minor discrepancies with a major discrepancy defined as one resulting in a change in diagnosis and treatment plans. We conducted patient follow-up via a retrospective review of the medical records to evaluate the clinical outcomes of the discrepancies. RESULTS: The rate of major discrepancies was 2.5%, and the rate of minor discrepancies was 11.4%. Major discrepancies led to a change in diagnosis or patient treatment plans, but did not lead to any increase in patient morbidity. CONCLUSION: The discrepancy rate at our institution was relatively insignificant, and patient care at the emergency department was not adversely affected by having radiology residents interpret CT and MRI scans after-hours and the attending radiologist review the interpretations the next morning. Still, further efforts are needed in order to reduce the frequency of major discrepancies.

A Serial Diffusion-weighted MR Imaging Study in Delayed Postanoxic Encephalopathy

BACKGROUND: Delayed postanoxic encephalopathy (DPE) is a demyelinating disorder in which cognitive, behavioral and movement deteriorations relapse in a few weeks after a complete recovery from initial hypoxic injury. To our knowledge, there have been no reports describing the temporal change of diffusion-weighted imaging (DWI) findings in DPE. METHODS: In a patient with DPE, serial brain MRI including DWI was performed at admission, 1, 3, 5 and 11 months after the onset of DPE. In order to obtain apparent diffusion coefficient (ADC) value of the lesion, we selected the same axial slices of the serial DWI and placed eight regions of interest (ROIs) in the periventricular white matter. Mean ADC values of ROIs and ADC map were serially compared. RESULTS: Compared to normal value, the ADC values of ROIs remained low at admission and one-month follow-up (0.68+/-0.08 and 0.67 +/- 0.08 X 10(-3)mm2/s) and then continued to increase on the 3, 5 and 11-month follow-up (0.78+/-0.05, 0.80+/-0.05 and 0.87 +/- 0.11 X 10(-3)mm2/s). Abnormal low signal density in both periventricular white matters on initial ADC map also disappeared slowly. Serial ADC maps revealed continuing diffuse cytotoxic edema over 1 month after the onset of DPE. CONCLUSIONS: Delayed cytotoxic edema associated with delayed clinical symptom of DPE might implicate that possible pathogenesis of DPE is apoptosis. DWI including ADC map might be useful to detect apoptotic brain lesion.

Non-ketotic Hyperglycemia Manifested as Occipital Lobe Complex Partial Status Epileticus

A 48-year-old woman presented with right visual aura evolving into versive and clonic seizures. The seizures were controlled not by anticonvulsants but bynormalization of blood sugar and not by anticonvulsants. T2-weighted MRI (T2WI) showed subcortical hypointensity and cortical hyperintensity with a low apparent diffusion coefficient (ADC) value in left occipital area, which reversed into a high ADC value at in follow-up images. Serial MRIs including diffu-sion- weighted images seemed to reveal tissue injury after partial status epilepticus and regional ischemia associated with non-ketotic hyperglycemia.

Regional MRI Volumetry and Longitudinal Regional Volume Curve for Detecting Hippocampal Sclerosis in Temporal Lobe Epilepsy

BACKGROUND: In order to improve MRI volumetry of hippocampus, we obtained the Longitudinal Regional Volume Curve (LRVC) of hippocampi and compared the sensitivity of LRVC with those of conventional hippocampal volume-tries (total and regional) in temporal lobe epilepsy (TLE). METHODS: Thirty-eight normal subjects and 24 TLE patients were included in this study. The pathology of all patients showed hippocampal sclerosis. The volume of the hippocampus was measured by a manual tracing in 3 mm-thickness coronal MRI slices perpendicular to the long axis of the hippocampus and a three-dimensional reconstruction. Total volume and regional volumes (anterior, middle, and posterior 1/3) of the right and left hippocampi were measured. Then, the focal hippocampal volume of each of the coronal slices (3 mm) was plotted in a X-Y graph to obtain LRVC. The presence and pattern of HA were determined in LRVC. RESULTS: The mean volume of right hippocampus (2512+/-629 mm3) was bigger than that of the left one (2262.6+/-563.2 mm3) in normal subjects. The normal range of right-left total volume difference was 3.6~495.2 mm3. The sensitivities of conventional volumetry, regional volumetry, and LRVC were 66.7%, 75%, and 83.3%, respectively. Eleven patients showed diffuse HA (11/20, 55.0%) and nine had focal HA (9/20, 45.0%). In focal HA, the middle and posterior HA were more frequent (6/9, 66.7%) than anterior HA. CONCLUSIONS: LRVC improved the sensitivity of MRI volumetry in detecting hippocampal sclerosis and could reveal the pattern (diffuse or focal) of HA. (J Korean Neurol Assoc 19(2):125~131, 2001)

Regional MRI Volumetry and Longitudinal Regional Volume Curve for Detecting Hippocampal Sclerosis in Temporal Lobe Epilepsy

BACKGROUND: In order to improve MRI volumetry of hippocampus, we obtained the Longitudinal Regional Volume Curve (LRVC) of hippocampi and compared the sensitivity of LRVC with those of conventional hippocampal volume-tries (total and regional) in temporal lobe epilepsy (TLE). METHODS: Thirty-eight normal subjects and 24 TLE patients were included in this study. The pathology of all patients showed hippocampal sclerosis. The volume of the hippocampus was measured by a manual tracing in 3 mm-thickness coronal MRI slices perpendicular to the long axis of the hippocampus and a three-dimensional reconstruction. Total volume and regional volumes (anterior, middle, and posterior 1/3) of the right and left hippocampi were measured. Then, the focal hippocampal volume of each of the coronal slices (3 mm) was plotted in a X-Y graph to obtain LRVC. The presence and pattern of HA were determined in LRVC. RESULTS: The mean volume of right hippocampus (2512+/-629 mm3) was bigger than that of the left one (2262.6+/-563.2 mm3) in normal subjects. The normal range of right-left total volume difference was 3.6~495.2 mm3. The sensitivities of conventional volumetry, regional volumetry, and LRVC were 66.7%, 75%, and 83.3%, respectively. Eleven patients showed diffuse HA (11/20, 55.0%) and nine had focal HA (9/20, 45.0%). In focal HA, the middle and posterior HA were more frequent (6/9, 66.7%) than anterior HA. CONCLUSIONS: LRVC improved the sensitivity of MRI volumetry in detecting hippocampal sclerosis and could reveal the pattern (diffuse or focal) of HA. (J Korean Neurol Assoc 19(2):125~131, 2001)

Comparison of Clinical Features and MRI Findings between Adamantinous and Papillary Craniopharyngioma / 대한내분비학회지

BACKGROUND: Craniopharyngioma is a suprasellar or intrasellar epithelial neoplasm that occurs in both children and adults. It accounts for 1.2 to 3 % of intracranial tumors with an incidence of 0.5 to 2 cases per one million populations each year. Recently, it has been postulated that it may have two pathogenetically separate subtypes, which are adamantinous and papillary craniopharyngioma, and that their clinical features may be different. However, there are some disagreements in this postulation. Therefore, we studied 22 consecutive patients with craniopharyngioma to evaluate the differences in clinical features and MRI findings between two subtypes. METHODS: We studied 22 patients with histologically proven craniopharyngioma after surgery at Samsung Medical center from 1995 to 1999. Thirteen patients were male, and nine patients were female. The average age was 30 years, with a range from 1 to 58 years. We divided 22 patients into two histopathologically separate subtypes; adamantinous and papillary subtypes. We compared the clinical features and MRI findings of two subtypes by reviewing medical records. RESULTS: Out of 22 patients with craniopharyngioma, 19 patients had an adamantinous subtype and 3 patients had a papillary subtype. The adamantinous subtype occurred frequently in the fifth decade and below twenty years, while the papillary subtype occurred predominantly in forth and fifth decades. The adamantinous subtype located in suprasellar or intrasellar portion as well as extrasellar portion, while the papillary subtype was restricted to the suprasellar location. The average tumor size of the adamantinous subtype was 3.7 cm, with a range from 1.4 to 6.0 cm, which was larger than that of the papillary subtype (average size 1.8 cm with a range from 1.5 to 2.3 cm, p< 0.05). The adamantinous subtype was predominantly cystic, while the papillary subtype was predominantly solid (p< 0.05). There were no significant differences in the preoperative clinical features and the postoperative complications between two subtypes. CONCLUSION: The adamantinous subtype had two peaks of occurrence in the fifth decade and below twenty years, while the papillary subtype occurred predominantly in forth and fifth decades. The adamantinous subtype was larger and had cystic portion, while the papillary subtype was smaller and had solid portion. The preoperative clinical features and the postoperative complications between two subtypes seemed not to be different.

Comparison of Immunohistochemical and Clinical Characteristics in Pituitary Adenoma with Acromegaly / 대한내분비학회지

BACKGROUND: It assumed that plurihormonal pituitary adenomas in acromegaly, which were immunohistochemically stained with other pituitary hormones in addition to GH and prolactin, would be originated from poorly differentiated cells. Therefore, we speculated that they might have higher growth rates and worse prognosis than monohormonal adenomas. To verify this speculation, we analyzed the frequency of plurihormonal adenomas and compared the clinical parameters and radiological invasiveness between plurihormonal adenoma and GH-prolactin adenoma in acromegaly. METHODS: We studied 38 patients with acromegaly (22 males and 16 females, mean age 40.7 years) who were underwent surgical removal of pituitary adenomas by TSA from January 1995 to February 1998. We performed immunohistochemical staining in these tumors using avidinbiotin peroxidase complex method. An adenoma was considered as immunoreactive when above 50 percents of tumor cells were stained with anti-hormonal antibodies. Invasiveness of tumors were evaluated by preoperative MRI findings on the basis of Hardys classification. RESULTS: The frequencies of plurihormonal and GH-prolactin adenomas were 42% and 58%, respectively. Plurihormonal adenoma included an adenoma which was not stained with prolactin, but with GH and other hormones. Prolactin immunoreactivity was found in 97%(37/38) of the tumors. Immunoreactivities to FSH, ACTH, LH, and TSH were found in 37.8%, 13.1%, 2.6% and 2.7%, respectively. There were no significant differences in age, basal serum GH and IGF-1 concentrations between plurihormonal and GH-prolactin adenomas. There were also no significant differences in response to TRH & LH stimulation tests and somatostatin & bromocriptine suppression tests between two groups. There were no differences in radiological invasiveness between two groups (plurihormonal adenoma, grade I 2, grade II 3, grade III 7, grade IV 4; GH-prolactin adenoma, grade I 3, grade II 6, grade III 9, grade IV 4). CONCLUSION: Plurihormonal adenomas were 44% and immunoreactivity to prolactin was 97% in pituitary adenomas in acromegaly. There were no significant differences in clinical parameters and radiological invasiveness between plurihormonal and GH-prolactin adenomas in acromegaly.

Primary Central Nervous System Lymphomas; A Clinicopathologic Study of 18 Cases

The incidence of a primary central nervous system (CNS) lymphoma in western countries is about 1% of all the intracranial tumors and has increased 2.2% over the last decade. A similar pattern of increased frequency is observed in Korea. Although most CNS lymphomas in western countries are high grade tumors carrying poor prognosis, the clinicopathologic features of the Korean CNS lymphoma have not been well studied. We report clinicopathological features of 18 cases of histologically proven primary brain lymphoma. The mean age of the patients was 50 years and there was no sex difference. The clinical and radiological characteristics included multiple site of occurrence, infrequent extracranial spread, and frequent seeding via cerebrospinal fluid. No patients were immune-compromised host. Of 18 cases, 15 cases were of B-lineage and 2 cases were of T-lineage. According to REAL classification, there were 12 cases of diffuse large B cell lymphoma, two cases of B cell lymphomas of small lymphoid cell, and two cases of peripheral T cell lymphoma, unspecified. The remaining subtypes were not subclassified because of inadequate material. Pleomorphic cytologic features and necrosis of varying extent were frequent in the cases of diffuse large B-cell lymphoma. These results suggest that overall clinicopathologic features of primary malignant lymphomas of the central nervous system in Korea are similar to those of western countries.

Delayed Effect of Contrast Enhancement in Brain Tumors on MRI

PURPOSE: To evaluate the degree of contrast enhancement of intracranial tumors on delayed (6-8min.) MR imaging after administration of Gd-DTPA. MATERIALS AND METHODS: Both immediate and delayed post-contrast MR imagings were studied prospectively in 35 patients. with brain tumors (11 gliomas, 6 meningiomas, 4 neurinomas, 5 parencymal metastases, 5 hemangioblastoma, 4 others) at either 0.5 T or 2.0 T unit. After precontrast TI-, proton-density, and T2-weighted spin echo images were obtained, each patient underwent Tl-weighted imaging immediately following infusion of 0.1 mmol/Kg of Gd-DTPA. Subsequently, the second postcontrast Tl-weighted images were obtained with no additional injection of the contrast media. Time-interval between the postcontrast immediate and the delayed images was approximately 6-8 minutes. Degree of contrast enhancement of the lesions was assessed both visually and quantitatively. For quantitative study, contrast enhancement ratio(CER) of tumors was calculated in both immediate and delayed post-contrast images. RESULTS: There was stronger visual enhancement in 7 of 11 cases with gliomas and 3 of 5 cases with parenchymal metastasis on delayed images when compared with immediate images, whereas all 10 cases of extraaxial tumors(meningiomas and neurinomas) showed decreased enhancement on delayed images. Quantitatively, mean CERs of gliomas and metastases were higher on delayed study than on immediate study by 20. 5% and 49.2%, respectively. Extraaxial tumors showed decrease of CER on delayed study by 19.7% as compared with that of immediate study. Hemangioblastomas showed visually poor enhancement on delayed image in 4 cases and equal enhancement on both immediate and delayed images in remaining one case, and quantitatively demonstrated decrease of CER on delayed study by 15.4%. CONCLUSION: Since there was more conspicuous contrast enhancement in many intraaxial tumors such as gliomas and metastases on 6-8 minutes delayed post-contrast MR study, the delayed post-contrast study may be needed in some intraaxial tumors for their characterization, and may also be helpful for the differential diagnosis between intraaxial and extraaxial tumors.