Original Research
Neuroradiology/Head and Neck Imaging
November 20, 2012

Accuracy of On-Call Resident Interpretation of CT Angiography for Intracranial Aneurysm in Subarachnoid Hemorrhage


OBJECTIVE. The purpose of this article is to evaluate the accuracy of preliminary on-call radiology resident interpretation of CT angiography (CTA) compared with digital subtraction angiography (DSA) in detecting cerebral aneurysms in subarachnoid hemorrhage (SAH).
MATERIALS AND METHODS. A retrospective review compared resident interpretations of head CTA performed after hours for SAH to the results of DSA. The sensitivity and specificity of resident interpretations were classified on a per-patient and per-aneurysm basis. The accuracy of resident interpretations was also determined according to aneurysm location and number.
RESULTS. Between January 2007 and December 2009, 83 patients with SAH underwent both CTA and DSA. DSA documented an aneurysm in 53 of 83 patients. Per patient, residents identified at least one aneurysm in 46 of 53 patients (87%). Per aneurysm, resident sensitivity and specificity for detecting aneurysms of any size were 62% and 91%, respectively, which improved for aneurysms 3 mm or larger to 73% and 97%, respectively. The posterior communicating and intracranial internal carotid arteries were resident “blind spots,” with aneurysms 3 mm or larger detected with sensitivities of 33% and 50%, respectively. In contrast, anterior communicating artery aneurysms were correctly identified 95% of the time. In only 35% of cases with multiple aneurysms did residents correctly identify more than one aneurysm.
CONCLUSION. The sensitivity of on-call resident interpretation of CTA for aneurysms in SAH is lower than expected, with a potential for delay in diagnosis and management in a small number of patients. Focused training to carefully review apparent blind spots and the frequency of multiple aneurysms may reduce inaccuracies.
At many academic centers, radiology residents provide preliminary interpretations for emergent cross-sectional studies after hours, on weeknights and weekends. A comprehensive evaluation of these preliminary reports is an essential component of any departmental quality assurance program, because misinterpretations create the potential for negative patient outcomes as well as opportunities for resident education and improvement. Prior investigations evaluating preliminary resident interpretation of emergent cross-sectional studies have shown low discrepancy rates compared with final attending physician readings. For example, misinterpretation rates for emergency unenhanced head CT have ranged from 0.1% to 7%, depending on whether clinically “significant” misses are considered with “insignificant” misses [14]. Many of these studies found lower discrepancy rates for more senior residents, results that helped support Accreditation Council for Graduate Medical Education (ACGME) mandates restricting first-year radiology residents from the independent call pool [5].
More recently, investigators have begun evaluating resident performance with regard to more advanced neuroimaging, such as MRI and CT angiography (CTA). A recent study reported a 7.2% discrepancy rate among residents interpreting MRI independently after hours, with the most common misses representing acute stroke, aneurysms, or vascular occlusions [6]. Another study found an approximately 13.5% discrepancy rate for resident interpretations of head and neck CTA [7]. In most of these studies, no adverse clinical outcomes were detected as the result of resident misinterpretations.
Spontaneous subarachnoid hemorrhage (SAH) is a life-threatening complication of intracranial aneurysms that requires prompt diagnostic assessment for proper therapeutic intervention. At most institutions, CTA has now replaced digital subtraction angiography (DSA) as the initial diagnostic test, both because of the rapidity with which it can be performed and its high accuracy in this setting [812]. Although small aneurysms can be difficult to perceive on CTA, this technique has consistently been proven reliable in the evaluation of aneurysms 3–4 mm or larger, with a recent investigation of a 64-MDCT angiography technique suggesting 100% and 93.7% sensitivities for aneurysms 3 mm or larger and smaller than 3 mm, respectively [13]. However, the accuracy of resident interpretations of CTA for aneurysm in the setting of SAH has not been thoroughly studied. Although a recent study evaluated resident accuracy in interpreting head and neck CTA, not all examinations were performed in the setting of SAH, and final attending physician readings were used as the reference standard [7]. Our purpose was to evaluate the accuracy of preliminary on-call radiology resident interpretation of CTA compared with DSA in detecting cerebral aneurysms in the setting of SAH.

Materials and Methods

A retrospective review was performed of all head CTA examinations ordered after hours over a 3-year period (January 1, 2007, through December 31, 2009) by emergency department physicians at our institution to evaluate the cause of spontaneous SAH. For this study, a coauthor used the PACS of our institution to identify all after-hours CTA studies performed for evaluation of spontaneous SAH. Studies were included in this investigation if a radiology resident provided a preliminary interpretation and if DSA was performed within 24 hours. Studies were excluded if the patient had a known aneurysm or history of treated aneurysm or if the resident indicated that the preliminary interpretation was aided by the “back-up” neuroradiology attending physician or fellow on call. Institutional review board approval was obtained for this retrospective study with a waiver of informed consent.
At our institution, radiology residents provide preliminary interpretations of emergency neuroradiology studies performed between 5 pm and 8 am Mondays through Fridays and all weekend long. During the week, a “dayfloat” resident covering the hours between 5 pm and 12 am is relieved by a “nightfloat” resident. Saturdays and Sundays are covered by residents as 12-hour shifts. Off-site neuroradiology fellows and attending physicians serve as backup if the resident needs aid in interpretation. Before being on call, residents have completed at least 1 month of a neuroradiology rotation, which includes CTA interpretation, and have also been exposed to CTA during morning didactic conferences. Preliminary interpretations are provided to clinicians and are available for review, because they become a permanent part of the medical record. Final interpretations are provided by attending neuroradiologists the following morning. Of note, DSA results are often available to attending physicians at the time of final CTA interpretation but are rarely available to residents at the time of preliminary interpretation. The majority of call time is taken between the residents’ second and third years, with first-year residents excluded from the independent call pool starting in July 2008 to comply with ACGME guidelines.
All CTA examinations were performed on a 64-MDCT scanner (LightSpeed VCT, GE Healthcare) available in our institution’s emergency department. Patients were scanned at a collimation of 64 × 0.625, pitch of 0.52, rotation time of 0.5 second, matrix of 512 × 512, thickness of 1.3 mm, and 140 kV. CTA images were obtained after the injection of 80 mL of ioversol (Optiray 320, Covidien) at 5 mL/s via an antecubital vein. Images were viewed on a PACS work-station as source images with axial, coronal, and sagittal maximum intensity projections (MIPs) with a thickness of 3 mm and overlap of 1 mm. Of note, because it may take 1 hour or longer for our 3D imaging laboratory to generate the curved multiplanar and 3D volume-rendered reformations, these images are usually not available to the residents at the time of preliminary interpretation.
DSA was performed with a biplane unit with rotational and 3D capabilities (Axiom Artis, Siemens Healthcare). Bilateral selective carotid and vertebral artery injections were performed, as required, with images obtained in anteroposterior, lateral, and oblique projections. Three-dimensional rotational DSA was also routinely performed, with MIP and shaded-surface-display images created from the rotational datasets. All DSA examinations were performed and interpreted by an attending physician experienced in neurovascular procedures.
The resident preliminary CTA interpretation and final DSA results were evaluated to determine the rate of agreement, including true-positives, true-negatives, false-positives, and false-negatives on a per-patient and per-aneurysm basis. Data were stratified according to aneurysm size (< 3 vs ≥ 3 mm), aneurysm location, and resident level of training (first through fourth year). Locations included the intracranial internal carotid artery (ICA), anterior communicating artery (AComA), middle cerebral artery (MCA), posterior communicating artery (PComA), and posterior circulation (vertebral and basilar arteries and their branches). The ICA segments evaluated in this study included segments C4 through C7, as defined by Bouthillier et al. [14]. The rate of agreement for identifying additional aneurysms in cases with more than one aneurysm was also calculated. For statistical analysis, the z test to compare two proportions was used to analyze the data, with a p value of less than 0.05 used to determine clinical significance.


Between January 1, 2007, and December 31, 2009, a total of 83 DSA studies followed an after-hours CTA for which a radiology resident provided a preliminary interpretation. DSA documented at least one aneurysm in 53 cases, and 30 cases were negative by angiography.

Accuracy of Resident Interpretation per Patient

Overall, residents correctly identified at least one aneurysm in 46 of the 53 (87%) CTA-positive studies, but incorrectly described seven (13%) CTA-positive studies as negative. Residents missed eight aneurysms in these seven studies, with missed aneurysms ranging from 2 to 7 mm in size. Table 1 shows how often residents correctly identified negative and positive CTA studies. From these data, the overall per-patient resident accuracy of 90% was calculated. Excluding first- and fourth-year residents (for whom too few studies met the inclusion criteria), second-year residents had the highest sensitivity at 93%; however, this was not significantly different compared with the 81% sensitivity for third-year residents (z = 1.201, z test to compare two proportions, one-tailed). Although positive preliminary interpretations were often accurate (positive predictive value, 98%), negative preliminary interpretations were less reliable (negative predictive value, 81%).

Accuracy of Resident Interpretation per Aneurysm

Overall, DSA identified 84 individual aneurysms in the 53 positive studies. Table 2 shows how accurate residents were in identifying individual aneurysms, both overall and stratified by aneurysm size (< 3 vs ≥ 3 mm). Although residents had an overall sensitivity of 62%, this improved to 73% for aneurysms equal to or greater than 3 mm. The difference in resident sensitivity for aneurysms less than 3 mm (29%) versus those equal to or greater than 3 mm was statistically significant (z = 3.373, z test to compare two proportions, one-tailed). Overall, resident accuracy on a per-aneurysm basis was 70%, improving to 81% for aneurysms larger than 3 mm.
TABLE 1: Resident Accuracy in Identifying at Least One Aneurysm on CT Angiography, on a Per-Patient Basis, Overall and by Level of Training
TABLE 2: Resident Accuracy in Identifying Individual Aneurysms, on a Per-Aneurysm Basis, Overall and by Aneurysm Size

Accuracy of Resident Interpretation With Multiple Aneurysms

Table 3 depicts how often residents were able to identify more than one aneurysm in the 17 cases for which DSA identified multiple aneurysms. Overall resident sensitivity was 35%, and although third-year residents tended to detect additional aneurysms more accurately than did second-year residents, this difference was not statistically significant (z = 0.696, z test to compare two proportions, one-tailed). Resident accuracy for identifying multiple aneurysms was 86%. As an example, Figures 1A and 1B illustrates a case in which the resident identified an 11-mm basilar artery aneurysm but missed a 5-mm left ICA ophthalmic segment (Bouthillier segment 6) aneurysm.

Accuracy of Resident Interpretation by Aneurysm Location

Table 4 depicts how accurate residents were in identifying aneurysms 3 mm or larger by location. Residents were most sensitive in identifying AComA aneurysms (95%), compared with lower sensitivities for MCA (76%), ICA (50%), posterior circulation (67%), and PComA (33%) aneurysms. Accuracy calculations also reflect this finding, because residents were 98% accurate in identifying AComA aneurysms but 84% accurate in identifying PComA aneurysms. Figures 2A and 2B shows a correctly identified 3-mm AComA aneurysm, a location for which residents were highly sensitive. On the other hand, Figures 3A, 3B, and 3C shows a 7-mm missed right PComA aneurysm, a location with which residents consistently had difficulty. Resident sensitivity for both AComA (z = 3.203) and MCA (z = 1.727) aneurysms were significantly greater than for PComA aneurysms.


The results presented here show that the sensitivity of on-call radiology resident interpretation of CTA for aneurysms in SAH is lower than expected. Overall, residents identified only 73% of aneurysms 3 mm or larger in size, and in 13% of patients, they failed to identify any aneurysm when at least one existed. The results also show that residents had a poor sensitivity for identifying additional aneurysms after locating an initial aneurysm. Finally, residents tended to miss more aneurysms in certain specific locations, such as the PComA and intracranial ICA, suggesting that these locations are potential resident “blind spots.” Because there are potential negative clinical ramifications for these misses, causes of and solutions for such discrepancies need to be considered.
Because the vast majority of patients who present to our emergency department with SAH are admitted to the neurosurgery service and are evaluated with conventional diagnostic angiography, regardless of the CTA results, opportunities for negative patient outcomes related to overnight resident misinterpretations are relatively few. However, the timely identification of an aneurysm in this clinical setting certainly allows the neurosurgical team to prepare for the most appropriate treatment approach (i.e., surgical vs endovascular) and directs neurointerventionalists to a target for therapy (because final neuroradiology attending physician interpretations may not be available at the time of angiography). Furthermore, at institutions where neuroangiography is not routinely available, a missed aneurysm may add an additional layer of delay to obtaining a DSA. Finally, resident aneurysm misinterpretation rates in the setting of SAH, when clinical suspicion for aneurysm is high, could be generalized to CTA performed in other emergency settings (i.e., trauma or headache), when a negative preliminary misinterpretation may result in a patient discharge home and a more prolonged delay in management. These potential negative clinical outcomes highlight the importance of determining why residents miss certain aneurysms and finding solutions to improve their sensitivity.
Technical factors should always be considered when analyzing why radiologic misinterpretations are made. The accuracy of CTA for aneurysm has been documented in many prior studies, with sensitivities for aneurysms 3 mm or larger consistently superseding 90% [1013, 15]. In fact, prior authors have touted this technique as the primary diagnostic modality in the selection of patients for surgical or endovascular treatment of ruptured aneurysms, with diagnostic DSA limited to patients with negative or inconclusive CTA findings [16]. Its utility in excluding aneurysm has been shown, with negative predictive values at or above 90% in many series [10, 13, 16]. Our investigation suggests a lower resident sensitivity for individual aneurysms in the setting of SAH than that reported for the CTA technique in the literature, as well as a lower overall negative predictive value. Although one expects sensitivity for aneurysms smaller than 3 mm to be somewhat lower, resident sensitivity for these small aneurysms (29%) is far below even that reported in the literature for 64-MDCT [13, 15, 1719].
TABLE 3: Resident Accuracy in Identifying Multiple Aneurysms on CT Angiography, Overall and by Level of Training
Fig. 1A 63-year-old woman with subarachnoid hemorrhage.
A, Emergency CT angiography axial source image (A) and maximum intensity projection (B) show 11-mm basilar artery (white arrow) and 5-mm left ophthalmic artery (black arrow) origin internal carotid artery (ICA) aneurysms. Although resident identified basilar artery aneurysm, left ICA aneurysm was missed.
Fig. 1B 63-year-old woman with subarachnoid hemorrhage.
B, Emergency CT angiography axial source image (A) and maximum intensity projection (B) show 11-mm basilar artery (white arrow) and 5-mm left ophthalmic artery (black arrow) origin internal carotid artery (ICA) aneurysms. Although resident identified basilar artery aneurysm, left ICA aneurysm was missed.
Although no formal evaluation of the accuracy of our 64-MDCT angiography technique has been performed, it should be noted that the CTA examinations in this study were consistently of high quality, and a review of the cases in this series found that all aneurysms identified on DSA were retrospectively visualized on CTA using raw data and MIPs. One limitation the residents did face in this study was lack of access to 3D reformatted images at the time of preliminary interpretation, including volume-rendered and curved reformatted images. Because reformatted images are thought to facilitate the evaluation of intracranial aneurysms [19], their availability should result in improved resident detection rates. However, this would require having someone such as a trained technologist or trainee available to obtain these reformatted images, because time constraints preclude the on-call residents from performing this task. Additionally, it should be noted that many radiologists, such as overnight teleradiologists and those working in the community hospital setting, often do not have reformatted images available and rely on their skills of interpreting raw CTA data and MIPs in detecting even small aneurysms.
Second, inadequacies in resident training should also be considered when determining causes of and solutions for resident misinterpretations. Although no statistically significant differences separated the accuracy of second- and third-year residents, this study shows that opportunities exist to increase aneurysm detection rates by improving resident education and exposure to CTA. In our series, residents tended to miss aneurysms in a few characteristic vascular distributions, such as the intracranial ICA, PComA, and posterior circulation (overall sensitivity, 50%), with improved detection rates in the AComA and MCA territories (overall sensitivity, 87%). As has been noted by prior investigators, adjacent high-attenuation bony structures have led to disproportionately high miss rates in the ICA and PComA territories [20]. Our experience suggests that most of the missed PComA aneurysms were obvious on the sagittal MIP images but were difficult, if not impossible, to detect on source images or coronal or axial MIP images. Instructing residents of this pitfall, and directing them to pay closer attention to these distributions, will likely improve detection rates.
Furthermore, residents incorrectly identified aneurysms that were not confirmed on DSA in three patients, two of whom had an entirely negative DSA for aneurysm. These false-positives also create an important opportunity for learning, because the incorrect identification of an aneurysm on CTA can subject the patient to unnecessary follow-up and more-invasive procedures. Strategies to help residents avoid some of the more common false-positive pitfalls, such as mistaking infundibula or tight vascular loops for aneurysms [11], can be used to minimize such overcalls.
Although there was no significant difference among second- and third-year residents in their identification of positive cases or individual aneurysms, third-year residents did identify a higher percentage of additional aneurysms in cases with multiple aneurysms. Although the largest aneurysm was usually identified in these cases, this was not always true, and in a few cases a second missed aneurysm was also treated. These results suggest that senior residents may be less susceptible to the “satisfaction of search” pitfall, perhaps because of a heightened understanding of how common additional aneurysms are. By stressing to residents the frequency of additional aneurysms, which ranges from 15% to 35% in the setting of SAH [21] and the importance of continuing to search, overall sensitivity is sure to improve.
TABLE 4: Accuracy of Resident Identification of Aneurysms ≥ 3 mm, by Location
Fig. 2A 69-year-old man with subarachnoid hemorrhage.
A, Coronal reformation from CT angiography examination shows 3-mm anterior communicating artery (AComA) aneurysm (arrow), which resident correctly identified.
Fig. 2B 69-year-old man with subarachnoid hemorrhage.
B, Frontal oblique digital subtraction angiography projection also shows AComA aneurysm (arrow).
Fig. 3A 65-year-old woman with subarachnoid hemorrhage.
A, Axial maximum intensity projection (MIP) shows 7-mm right posterior communicating artery (PComA) aneurysm (arrow), which was missed by resident.
Fig. 3B 65-year-old woman with subarachnoid hemorrhage.
B, Sagittal MIP of CT angiography (B) and lateral digital subtraction angiography projection (C) also show right PComA aneurysm (arrow). This was only aneurysm in this patient.
Fig. 3C 65-year-old woman with subarachnoid hemorrhage.
C, Sagittal MIP of CT angiography (B) and lateral digital subtraction angiography projection (C) also show right PComA aneurysm (arrow). This was only aneurysm in this patient.
Other factors that affect resident accuracy are difficult to counteract, including the busy and distracting emergency department setting in which interpretations are often being rendered. Additionally, long overnight shifts spanning many consecutive evenings produce excessive fatigue, which also can affect accuracy. Such factors affect residents throughout the country and are starting to be dealt with on a national level [22].
One potential limitation of this study is the relatively small sample size, with interpretations made by a single pool of residents at one academic institution. Although this could raise concern over how generalizable our results are, it should be noted that ours is a large training program with 48 different residents in the call pool over the 3 years these emergency CTA examinations were performed, with the vast majority of our residents passing the American Board of Radiology oral examination with the first attempt. Another potential limitation is the fact that our residents did not have 3D images available at the time of preliminary interpretation, considering that a previous study [10] indicated that their residents had 3D images available. However, it should be noted that having 3D and curved reformatted images at the time of preliminary interpretations is unusual in most academic and private practice settings because it is quite difficult to provide 3D services around the clock. Therefore, we think that our evaluation best simulates the practice environment in which our residents will find themselves in the future. Finally, some may find it a limitation that we are comparing resident preliminary interpretation of CTA not to the attending physician’s interpretation of the same study but instead to the attending physician’s interpretation of a different modality (i.e., DSA). At our institution, although there is always a neuroradiology fellow and attending physician on call to assist residents if needed, final attending physician CTA interpretations may not be rendered until after DSA has been performed and can be influenced, and therefore biased, by these results. Additionally, as we have already stated, all aneurysms seen on DSA could be identified on CTA in retrospect. However, to counteract this limitation, future studies could compare resident preliminary CTA interpretations to blinded attending physician CTA interpretations.
In conclusion, our results show that the sensitivity of resident interpretation of CTA for aneurysm is lower than expected, though there are educational opportunities to prevent certain characteristic misinterpreta tions. By empowering trainees with an understanding of the most common errors that lead to resident misinterpretations, aneurysm detection rates should approach what is expected of the MDCT angiography technique and lead to more timely and accurate treatment planning.


Erly WK, Berger WG, Krupinski E, Seeger JF, Guisto JA. Radiology resident evaluation of head CT scan orders in the emergency department. AJNR 2002; 23:103–107
Lal NR, Murray UW, Eldevik P, Desmond JS. Clinical consequences of misinterpretations of neuroradiologic CT scans by on-call radiology residents. AJNR 2000; 21:124–129
Wysoki MG, Nassar CJ, Koenigsberg R A, Novelline RA, Faro SH, Faerber FN. Head trauma: CT scan interpretation by radiology residents versus staff. Radiology 1998; 208:125–128
Strub WM, Vagal AA, Tomsick T, Moulon JS. Overnight resident preliminary interpretations on CT examinations: should the process continue? Emerg Radiol 2006; 13:19–23
Amis ES. New program requirements for diagnostic radiology: update and discussion of the more complex requirements. AJR 2008; 190:2–4
Filippi CG, Schneider B, Burbank HN, Alosfrom GF, Linnell G, Ratkovits B. Discrepancy rates of radiology resident interpretations of on-call neuroradiology MR imaging studies. Radiology 2008; 249:972–979
Meyer RE, Nickerson JP, Burbank HN, Alsofrom GF, Linnell GJ, Filippi CG. Discrepancy rates of on-call radiology residents’ interpretations of CT angiography studies of the neck and circle of Willis. AJR 2009; 193:527–532
Colen TW, Wang LC, Ghodke BV, Cohen WA, Hollingworth W, Anzai Y. Effectiveness of MDCT angiography for the detection of intracranial aneurysms in patients with nontraumatic subarachnoid hemorrhage. AJR 2007; 189:898–903
Lubicz B, Levivier M, Francois O, et al. Sixty-four-row multisection CT angiography for detection and evaluation of ruptured intracranial aneurysms: interobserver and intertechnique reproducibility. AJNR 2007; 28:1949–1955
Yoon DY, Lim KJ, Choi CS, Cho BM, Oh SM, Chang SK. Detection and characterization of intracranial aneurysms with 16-channel multidetector row CT angiography: a prospective comparison of volume-rendered images and digital subtraction angiography. AJNR 2007; 28:60–67
Westerlaan HE, van Dijk JMC, Jansen-van der Weide MC, et al. Intracranial aneurysms in patients with subarachnoid hemorrhage: CT angiography as a primary examination tool for diagnosis—systematic review and meta-analysis. Radiology 2011; 258:134–145
Chang LK, Liew NS, Soh HL, Tan SZ, Wong SH. Clinical utility of 64-row multislice CT angiography in the detection of cerebral aneurysms in acute subarachnoid haemorrhage. Med J Malaysia 2008; 63:131–136
Li Q, Lv F, Li Y, Luo T, Li K, Xe P. Evaluation of 64-section CT angiography for detection and treatment planning of intracranial aneurysms by using DSA and surgical findings. Radiology 2009; 252:808–815
Bouthillier A, van Lovernan HR, Keller JT. Segments of the internal carotid artery: a new classification. Neurosurgery 1996; 38:425–433
Korogi Y, Takahashi M, Katada K, et al. Intracranial aneurysms: detection with three-dimensional CT angiography with volume rendering—comparison with conventional angiographic and surgical findings. Radiology 1999; 211:497–506
Westerlaan HE, Gravendeel J, Fiore D, et al. Multislice CT angiography in the selection of patients with ruptured intracranial aneurysms suitable for clipping or coiling. Neuroradiology 2007; 49:997–1007
McKinney AM, Palmer CS, Truwit CL, Karagulle A, Teksam M. Detection of aneurysms by 64-section multidetector CT angiography in patients acutely suspected of having an intracranial aneurysm and comparison with digital subtraction and 3D rotational angiography. AJNR 2008; 29:594–602
Teksam M, Mckinney A, Casey S, Asis M, Keiffer S, Truwit CI. Multi-section CT angiography for detection of cerebral aneurysms. AJNR 2004; 25:1485–1492
Villablanca JP, Reza J, Hooshi P, et al. Detection and characterization of very small cerebral aneurysms by using 2D and 3D helical CT angiography. AJNR 2002; 23:1187–1198
Sakamoto S, Kiura Y, Shibukawa M, Ohba S, Arita K, Kurisu K. Subtracted 3D CT angiography for evaluation of internal carotid artery aneurysms: comparison with conventional digital subtraction angiography. AJNR 2006; 27:1332–1337
Kaminogo M, Yonekura M, Shibata S. Incidence and outcome of multiple intracranial aneurysms in a defined population. Stroke 2003; 34:16–21
Mainiero MB, Davis LP, Chertoff JD. Resident duty hour limits: recommendations by the IOM and the response from the radiology community. J Am Coll Radiol 2010; 7:56–60

Information & Authors


Published In

American Journal of Roentgenology
Pages: 1436 - 1441
PubMed: 22109300


Submitted: March 1, 2011
Accepted: April 25, 2011


  1. CT angiography
  2. intracranial aneurysm
  3. radiology resident interpretation
  4. subarachnoid hemorrhage



Aaron R. Hochberg
Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02115.
Department of Radiology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA.
Rafael Rojas
Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02115.
Ajith J. Thomas
Department of Neurosurgery, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA.
Arra Suresh Reddy
Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02115.
Rafeeque A. Bhadelia
Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Ave, Boston, MA 02115.


Address correspondence to A. R. Hochberg ([email protected]).

Metrics & Citations



Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options


View PDF

PDF Download

Download PDF







Copy the content Link

Share on social media