Multilevel Strip Pooling-Based Convolutional Neural Network for the Classification of Carotid Plaque Echogenicity.

Carotid plaque echogenicity in ultrasound images has been found to be closely correlated with the risk of stroke in atherosclerotic patients. The automatic and accurate classification of carotid plaque echogenicity is of great significance for clinically estimating the stability of carotid plaques and predicting cardiovascular events. Existing convolutional neural networks (CNNs) can provide an automatic carotid plaque echogenicity classification; however, they require a fixed-size input image, while the carotid plaques are of varying sizes. Although cropping and scaling the input carotid plaque images is promising, it will cause content loss or distortion and hence reduce the classification accuracy. In this study, we redesign the spatial pyramid pooling (SPP) and propose multilevel strip pooling (MSP) for the automatic and accurate classification of carotid plaque echogenicity in the longitudinal section. The proposed MSP module can accept arbitrarily sized carotid plaques as input and capture a long-range informative context to improve the accuracy of classification. In our experiments, we implement an MSP-based CNN by using the visual geometry group (VGG) network as the backbone. A total of 1463 carotid plaques (335 echo-rich plaques, 405 intermediate plaques, and 723 echolucent plaques) were collected from Zhongnan Hospital of Wuhan University. The 5-fold cross-validation results show that the proposed MSP-based VGGNet achieves a sensitivity of 92.1%, specificity of 95.6%, accuracy of 92.1%, and F1-score of 92.1%. These results demonstrate that our approach provides a way to enhance the applicability of CNN by enabling the acceptance of arbitrary input sizes and improving the classification accuracy of carotid plaque echogenicity, which has a great potential for an efficient and objective risk assessment of carotid plaques in the clinic.

A decade of improvement in the management of New Zealand ST-elevation myocardial infarction (STEMI) patients: results from the New Zealand Acute Coronary Syndrome (ACS) Audit Group national audits of 2002, 2007 and 2012.

AIMS: To audit the management of ST-segment elevation myocardial infarction (STEMI) patients admitted to a New Zealand Hospital over three 14-day periods to review their number, characteristics, management and outcome changes over a decade. METHODS: The acute coronary syndrome (ACS) audits were conducted over 14 days in May of 2002, 2007 and 2012 at New Zealand Hospitals admitting patients with a suspected or definite ACS. Longitudinal analyses of the STEMI subgroup are reported. RESULTS: From 2002 to 2012, the largest change in management was the proportion of patients undergoing reperfusion by primary PCI from 3% to 15% and 41%; P<0.001, and the rates of second antiplatelet agent use in addition to aspirin from 14% to 62% and 98%; P<0.001. The use of proven secondary prevention medications at discharge also increased during the decade. There were also significant increases in cardiac investigations for patients, especially echocardiograms (35%, 62% and 70%, P<0.001) and invasive coronary angiograms (31%, 58% and 87%, P<0.001). Notably even in 2012, one in four patients presenting with STEMI did not receive any reperfusion therapy. CONCLUSIONS: Substantial improvements have been seen in the management of STEMI patients in New Zealand over the last decade, in accordance with evidenced-based guideline recommendations. However, there appears to be considerable room to optimise management, particularly with the use of timely reperfusion therapy for more patients.

Vascular imaging before intravenous thrombolysis: consequences of in-hospital delay in applying two diagnostic procedures.

BACKGROUND: Vascular imaging is increasingly used for diagnosis of arterial occlusions in acute ischemic stroke (AIS) patients. Our aim was to determine whether computed tomography angiography (CTA) and Doppler/duplex ultrasound (DUS) before intravenous thrombolysis (IVT) is associated with a delay in time-to-treatment. METHODS: Observational analysis of a prospective cohort of AIS patients treated with IVT from January 2009 to December 2012. Patients were classified into three groups: the noncontrast computed tomography (NCCT) group (patients studied only with NCCT before IVT), CTA group (patients who underwent CTA in addition to NCCT), and the DUS group (patients studied with NCCT+DUS). RESULTS: We treated 244 patients: 116 patients (47.5%) were studied with NCCT, 79 (32.4%) with CTA, and 49 (20.1%) with DUS. Door-to-needle time was significantly higher in the CTA group (median 60 [48-77] minutes) than in the NCCT group (51.5 [40-65]) and DUS group (48 [42-61]) (P = .008). No differences were observed for onset-to-door time and onset-to-needle time. In the multivariate linear regression analysis, onset-to-door time, prehospital stroke code activation, and performance of CTA influenced door-to-needle time. CONCLUSIONS: Performing CTA before IVT seems to increase door-to-needle time. Vascular imaging based on DUS should be considered only if this does not lead to in-hospital delay.

Limitations of routine neuroimaging in predicting outcomes of preterm infants.

INTRODUCTION: Preterm births are increasing in number and while the rates of cerebral palsy have declined, there are increasing numbers of infants who survive with handicaps. In some studies, up to 50 % of children will have morbidity when followed up to school age. METHODS: A review of current literature was conducted to determine the validity of routine cranial ultrasound scans (CUS) to predict neurodevelopmental outcomes, including motor and cognitive deficits. We also reviewed the additional benefit offered by including MRI scans in scanning protocols to enhance the reliability in predicting the neurodevelopmental sequelae of prematurity. RESULTS: CUS is valuable as a screening tool to determine significant brain injury when conducted regularly over the first weeks of life in preterm infants. Subtle changes on CUS are difficult to interpret and more precise information is offered by performing MRI scans. These are most often carried out at term equivalent age but earlier scans may be just as useful in predicting neurocognitive outcomes. When MRI scans are either normal or seriously abnormal, there is a very clear correlation with outcome to 2 years of age. Mild and moderate degrees of injury defined on MRI need more sophisticated scanning sequences to determine the likelihood of associated sequelae. Follow-up to school age is essential to diagnose more subtle cognitive delays. CONCLUSION: CUS provides a good screening tool to detect serious brain injury resulting in motor handicaps but MRI scans are complementary and necessary to accurately predict the outcomes of preterm infants, especially cognitive delays.

Imaging the premature brain: ultrasound or MRI?

Neuroimaging of preterm infants has become part of routine clinical care, but the question is often raised on how often cranial ultrasound should be done and whether every high risk preterm infant should at least have one MRI during the neonatal period. An increasing number of centres perform an MRI either at discharge or around term equivalent age, and a few centres have access to a magnet in or adjacent to the neonatal intensive care unit and are doing sequential MRIs. In this review, we try to discuss when best to perform these two neuroimaging techniques and the additional information each technique may provide.

The significance of the ultrasound diagnostics in evaluation of the emboligenic pathogenesis of transient ischemic attacks.

The objective of this study was to examine the possibilities of ultrasound diagnostics in the evaluation of emboligenic pathogenesis of transient ischemic attacks (TIAs) and the frequency of specific origins of embolism. A total of 150 adult patients with TIA and 50 control patients, were examined by neurosonologic, echocardiographic and venous ultrasound examination. Emboligenic pathogenesis of TIA was established in 36.6% of patients. Microembolic signals were detected in 22.7% of the whole group, and 61.8% in emboligenic TIA subgroup. Artery-to-artery embolism from ulcerated plaque of the carotid arteries was found in 12.6% of patients, from the aortic arch atheroma in 3.3% and cardioembolism in 12.6% (atrial fibrillation 7.3%, atrial septal aneurysm 2%, mitral valve prolapse 2%, mechanical heart valve 0.7%, left atrium thrombus 0.7%). Paradoxic embolism with the patent foramen ovale was established in 6% of patients, and with the pulmonary right-to-left shunt in 2%. Correlation with controls showed significantly higher frequency of the ulcerated carotid plaque and frequency of microembolic signals in the TIA group (p < 0.05). The patients with potential sources of embolism had a greater risk of developing TIA than those without these sources.

Diagnostic approach to children with minor traumatic brain injury.

STUDY PURPOSE: To analyse the management of minor traumatic brain injury (MTBI) in paediatric hospitals in Germany. METHODS: An electronic survey was sent to 72 children hospitals. RESULTS: All participating (45/72; 62.5 %) hospitals had facilities to perform an electroencephalogram (EEG), 98 % cranial ultrasonography, 94 % MRI studies, and 87 % a CT scan. The initial Glasgow Coma Scale, the clinical presentation/neurological deficits, the intensity of the trauma and external/visible injuries were most important for initial assessment. The main reason for in-patient monitoring was initial clinical neurologic presentation (44 %). X-ray scans were used routinely in only 2.2 %, cMRI scans in 6.7 % and cCT scans in 13.3 %; approximately one third employed ultrasonography. In 22.2 % was an EEG part of the routine diagnostic work-up. Inpatient monitoring for 24-48 h was done in 80 %. CONCLUSIONS: Children with MTBI are often monitored clinically without resorting to potentially harmful and expensive diagnostic procedures (cCT scans).

Long-term prognostic value of the right ventricular myocardial performance index compared to other indexes of right ventricular function in patients with moderate chronic heart failure.

BACKGROUND: The ventricular myocardial performance index (MPI) is a feasible echocardiographic parameter for the evaluation of patients with chronic heart failure (CHF). The long-term prognostic role of right ventricular MPI (RV MPI) has been already assessed in patients with more advanced CHF but data are lacking in moderate CHF. The aim of the study is to evaluate the possible prognostic role of RV MPI in moderate CHF patients compared to others traditional RV parameters. METHODS: From 2003 to 2004 we enrolled 95 consecutive NYHA class II CHF patients (65 males and 30 females), with the mean age of 66 ± 11 years with left ventricular ejection fraction (LVEF) <40%, on optimal medical treatment. All patients were evaluated clinically and by echocardiography with a follow-up of 5 years (combined end point: cardiovascular mortality and hospitalization for HF). RESULTS: RV MPI was 0.45 ± 0.36, tricuspid annular plane systolic excursion was 21 ± 8 mm, RV fractional area change was 42 ± 12%, systolic pulmonary artery pressure was 33 ± 9 mmHg, and acceleration time of pulmonic flow was 115.5 + 22.62 msec. After the 5 year follow-up the total mortality was 24.2% and HF hospitalization rate was 33%. At Cox multivariate analysis only an RV MPI superior to median value (>0.38) and tricuspid annular plane systolic excursion inferior to median value (<18 mm) had shown a significant prognostic role. CONCLUSION: The RV MPI in a population of moderate CHF showed to have a more long-term powerful prognostic value than other conventional and traditional echocardiographic right ventricular functional parameters.

The role of intraoperative micro-Doppler ultrasound in verifying proper clip placement in intracranial aneurysm surgery.

INTRODUCTION: Aneurysmal subarachnoid hemorrhage constitutes a clinical entity associated with high mortality and morbidity. It is widely accepted that improper clip placement may have as a result of incomplete aneurysm occlusion and/or partial or complete obstruction of an adjacent vessel. Various modalities, including intraoperative or postoperative digital subtracting angiography, near-infrared indocyanine green angiography, micro-Doppler ultrasonography (MDU), and neurophysiological studies, have been utilized for verifying proper clip placement. The aim of our study was to review the role of MDU during aneurysmal surgery. METHODS: A literature search was performed using any possible combination of the following terms: "aneurysm," "brain," "cerebral," "clip," "clipping," "clip malpositioning," "clip repositioning," "clip suboptimal positioning," "Doppler," "intracranial," "microsurgery," "micro-Doppler," "residual neck," "ultrasonography," "ultrasound," and "vessel occlusion". Additionally, reference lists from the retrieved articles were reviewed for identifying any additional articles. Case reports and miniseries were excluded. RESULTS: A total of 19 series employing intraoperative MDU during aneurysmal microsurgery were retrieved. All studies demonstrated that MDU accuracy is extremely high. The highest reported false-positive rate of MDU was 2 %, while the false-negative rate was reported as high as 1.6 %. It has been demonstrated that the presence of subarachnoid hemorrhage, specific anatomic locations, and large size may predispose to improper clip placement. Intraoperative MDU's technical limitations and weaknesses are adequately identified, in order to minimize the possibility of any misinterpretations. CONCLUSION: Intraoperative MDU constitutes a safe, accurate, and low cost imaging modality for evaluating blood flow and for verifying proper clip placement during microsurgical clipping.

Comparison of different reconstruction algorithms for three-dimensional ultrasound imaging in a neurosurgical setting.

BACKGROUND: Freehand three-dimensional ultrasound imaging (3D-US) is increasingly used in image-guided surgery. During image acquisition, a set of B-scans is acquired that is distributed in a non-parallel manner over the area of interest. Reconstructing these images into a regular array allows 3D visualization. However, the reconstruction process may introduce artefacts and may therefore reduce image quality. The aim of the study is to compare different algorithms with respect to image quality and diagnostic value for image guidance in neurosurgery. METHODS: 3D-US data sets were acquired during surgery of various intracerebral lesions using an integrated ultrasound-navigation device. They were stored for post-hoc evaluation. Five different reconstruction algorithms, a standard multiplanar reconstruction with interpolation (MPR), a pixel nearest neighbour method (PNN), a voxel nearest neighbour method (VNN) and two voxel based distance-weighted algorithms (VNN2 and DW) were tested with respect to image quality and artefact formation. The capability of the algorithm to fill gaps within the sample volume was investigated and a clinical evaluation with respect to the diagnostic value of the reconstructed images was performed. RESULTS: MPR was significantly worse than the other algorithms in filling gaps. In an image subtraction test, VNN2 and DW reliably reconstructed images even if large amounts of data were missing. However, the quality of the reconstruction improved, if data acquisition was performed in a structured manner. When evaluating the diagnostic value of reconstructed axial, sagittal and coronal views, VNN2 and DW were judged to be significantly better than MPR and VNN. CONCLUSION: VNN2 and DW could be identified as robust algorithms that generate reconstructed US images with a high diagnostic value. These algorithms improve the utility and reliability of 3D-US imaging during intraoperative navigation.

Cranial ultrasonography and transfontanellar Doppler in premature neonates (24-32 weeks of gestation): dynamic evolution and association with a severe adverse neurological outcome at hospital discharge in the Aquitaine cohort, 2003-2005.

OBJECTIVE: To correlate the short-term neurological outcome of early cerebral abnormalities using cranial ultrasonography (US) in premature newborns at their hospital discharge. METHODS: Each newborn born <33 weeks of gestational age (GA) included in a prospective cohort benefited of 3 US: two early in the first week of life (D3 and D8) and one later (Months 1-2) standardized US pulsed Doppler. A US abnormality was ≥one morphologic abnormality (moderate: intra-ventricular hemorrhage (IVH) grades 1-2; severe: IVH 3-4, periventricular leukomalacia, persistent flares). Correlates of having a severe adverse neurological outcome were analyzed using a stepwise backward logistic regression adjusted model with gestational age, early cerebral abnormality at Days 3-8, velocity and with variables with correlation probabilities with p<0.25 in the univariate analysis among occurring co-morbidity events previously defined. Two adjusted logistic regression analyses were conducted including or not velocity data because of missing information. RESULTS: Among 452 premature included, 11.3% did not have an early US, 74.8% had a normal early US, 13.9% ≥one early morphological US abnormality (10.0% moderate, 3.9% severe). At hospital discharge, 40% were still alive with a normal late US, 33% alive with ≥one late morphological US abnormality (10% moderate, 23% severe), and 10% died. Adjusted correlates of a late US severe abnormality or a neurological related death at hospital discharge were: early US abnormality (aOR: 8.7, 95% CI: 2.3-33.6), GA<29 weeks (aOR: 2.8 95% CI: 1.4-5.4). CONCLUSION: This study shows that early US morphological abnormalities increase significantly when the GA decreases and is highly predictive of the occurrence of a further late US severe abnormality or neurological related death at hospital discharge.

Predictors of hyperperfusion syndrome before and immediately after carotid artery stenting in single-photon emission computed tomography and transcranial color-coded real-time sonography studies.

BACKGROUND: Hyperperfusion syndrome (HPS) is a critical complication after carotid artery stenting (CAS) and carotid endarterectomy (CEA). OBJECTIVE: To identify predictors of HPS before and immediately after CAS. METHODS: We analyzed patients who underwent elective CAS from 2005 to 2008, and underwent single-photon emission computed tomography (SPECT) and transcranial color-coded real-time sonography before and immediately after CAS. HPS was defined as post-CAS deteriorating neurological conditions with headache not secondary to cerebral ischemia. We assessed the measures of blood flow between the two cortical hemispheres by taking the ratio of cerebral blood flow (CBF) of the affected to unaffected hemisphere excluding any ischemic/infarcted areas (asymmetry index); the measures of blood flow within each cortical hemisphere by comparing the CBF in the affected cortical hemispheric area to the CBF in the ipsilateral cerebellar hemisphere (middle cerebral artery [MCA]-to-cerebellar activity ratio); cerebral vasoreactivity (CVR); MCA mean blood flow velocity in the affected hemisphere; and MCA mean blood flow velocity ratio (preoperative to postoperative). RESULTS: Sixty-four patients were analyzed retrospectively. Nine patients presented with HPS. Logistic regression analysis showed that CVR (P<.01) and MCA mean blood flow velocity (P<.05) were the significant predictors among the pre-CAS variables, and that MCA mean blood flow velocity ratio (P<.05) and MCA-to-cerebellar activity ratio change (P<.05) were significant predictors among the post-CAS variables. CONCLUSION: SPECT and transcranial color-coded real-time sonography studies are useful in predicting HPS.

Reader variability in the use of diagnostic terms to describe white matter lesions seen on cranial scans of severely premature infants: the ELGAN study.

PURPOSE: To evaluate reader variability of white matter lesions seen on cranial sonographic scans of extreme low gestational age neonates (ELGANs). METHODS: In 1,452 ELGANs, cranial sonographic scans were obtained in the first and second postnatal weeks, and between the third postnatal week and term. All sets of scans were read independently by two sonologists. We reviewed the use of four diagnostic labels: early periventricular leucomalacia, cystic periventricular leucomalacia, periventricular hemorrhagic infarction (PVHI), and other white matter diagnosis, by 16 sonologists at 14 institutions. We evaluated the association of these labels with location and laterality of hyperechoic and hypoechoic lesions, location of intraventricular hemorrhage, and characteristics of ventricular enlargement. RESULTS: Experienced sonologists differed substantially in their application of the diagnostic labels. Three readers applied early periventricular leucomalacia to more than one fourth of all the scans they read, whereas eight applied this label to ≤5% of scans. Five applied PVHI to ≥10% of scans, while three applied this label to ≤5% of scans. More than one third of scans labeled cystic periventricular leucomalacia had unilateral hypoechoic lesions. White matter abnormalities in PVHI were more extensive than in periventricular leucomalacia and were more anteriorly located. Hypoechoic lesions on late scans tended to be in the same locations, regardless of the diagnostic label applied. CONCLUSIONS: Experienced sonologists differ considerably in their tendency to apply diagnostic labels for white matter lesions. This is due to lack of universally agreed-upon definitions. We recommend reducing this variability to improve the validity of large multicenter studies.

Fetal brain asymmetry: in utero sonographic study of normal fetuses.

OBJECTIVE: The objective of the study was to evaluate the magnitude of normal fetal brain asymmetry. STUDY DESIGN: This was a prospective study. Normal fetuses between 19-28 weeks of gestation were studied. The cerebral atria, occipital cortex, and hemispheres in both sides were measured. The difference between each side was evaluated and was correlated with sex, head biometry, and estimated weight. RESULTS: Four hundred six fetuses were studied. Mean atrial width was larger in the males and on the left side (5.2% and 6.5%, respectively). Mean cortical width was 2.6% larger in males but 5.5% thinner on the left side. Mean hemisphere width was larger in males and on the left side (2.3% and 1.5%, respectively). The atria and the cortex presented an inverse relationship regarding fetal growth parameters. CONCLUSION: Brain asymmetry represents normal fetal brain developmental phenomena. It is sex dependent and lateralized in most cases to the left. Lateralization was more accentuated in males.

Is routine TORCH screening warranted in neonates with lenticulostriate vasculopathy?

BACKGROUND: Congenital infections are associated with a wide spectrum of clinical symptoms, including lenticulostriate vasculopathy (LSV). OBJECTIVE: To determine the relationship between LSV and congenital infections, as diagnosed by TORCH serology and viral culture for cytomegalovirus (CMV). METHODS: All neonates with LSV admitted to our neonatal intensive-care unit from 2004 to 2008 were included in the study. Results of maternal and neonatal TORCH testing were evaluated. RESULTS: During the study period, cranial ultrasound scans were performed in 2,088 neonates. LSV was detected in 80 (4%) neonates. Maternal and/or neonatal serological TORCH tests were performed in 73% (58/80) of cases. None of the mothers or infants (0 of 58) had positive IgM titres for Toxoplasma, rubella, CMV or herpes simplex virus. Additional urine culture for CMV was performed in 38 neonates. None of the infants (0 of 38) had a positive CMV urine culture test. CONCLUSIONS: Routinely applied efforts to diagnose congenital infections in cases presenting with LSV have a poor yield. Routine TORCH screening in neonates with LSV cases should only be regarded as mandatory once well-designed studies demonstrate a clear diagnostic benefit.

Supratentorial abnormalities in the Chiari II malformation, III: The interhemispheric cyst.

OBJECTIVE: Our clinical observations noted an interhemispheric "cyst," a protrusion of the posterior superior third ventricle, in fetuses with myelomeningocele (MMC). The study analyzed the frequency of this observation and features that influence its visualization. METHODS: We searched for cases of fetal MMC sonographically detected between 1999 and 2007. Intracranial findings were retrospectively reviewed with attention to the interhemispheric cistern. Additionally, we reviewed 25 fetuses without a central nervous system anomaly and 10 fetuses with ventriculomegaly but no MMC. RESULTS: Among 89 fetuses identified, the mean gestational age was 22 weeks 4 days. Thirty-eight (43%) had an interhemispheric cyst. The frequency was similar on sonograms judged to be well visualized compared with studies judged to be suboptimally visualized. The degree of ventriculomegaly, timing of diagnosis, and severity of the Chiari II malformation did not appear to influence the frequency of the finding. Among fetuses without a central nervous system anomaly, no interhemispheric cysts were detected; a cyst was detected in 1 of 10 fetuses with ventriculomegaly. Interhemispheric cysts were more likely to be detected in fetuses with the Chiari II malformation than fetuses with ventriculomegaly but without the Chiari II malformation (P = .04). CONCLUSIONS: Interhemispheric cysts are a common supratentorial feature of the Chiari II malformation. Their presence appears to be unrelated to other features of the Chiari II malformation. Although interhemispheric cysts are seen in other abnormal fetuses, their striking prevalence in the Chiari II malformation should lead to a thorough examination for MMC.

Frontal bone window improves the ability of transcranial color-coded sonography to visualize the anterior cerebral artery of Asian patients with stroke.

The frontal bone window (FBW) is a promising approach in evaluating the anterior cerebral artery. The goal of the present study was to determine the rates of detection of the basal cerebral arteries by using the FBW alone and a combination of the FBW with the temporal bone window (TBW) in 163 patients. The combined application improved detection rates of A1 (58.6% versus 46.0%, P = .001) and A2 (43.6% versus 6.7%, P < .001) compared with the TBW alone.

Fetal cerebral ventricular pointing as a marker of spina bifida: incidence and observational agreement.

OBJECTIVE: Occipital cerebral ventricular pointing is a promising intracranial marker for spina bifida. We sought to determine (1) the incidence of pointing in fetuses with and without spina bifida and (2) inter-observer and intraobserver agreement for visualizing pointing. METHODS: Second-trimester transverse axial images of the lateral cerebral ventricles from an equal number of affected and unaffected fetuses were independently reviewed in a blinded fashion by 2 examiners. RESULTS: Sixty-two fetuses (31 with isolated spina bifida and 31 without) had sonographic examinations at 19.2+/-1.3 weeks (mean+/-SD). Pointing was present in 77.3% of fetuses with spina bifida and 15% of those without by the first examiner and in 78.3% of fetuses with spina bifida and 10% of those without by the second examiner. Interobserver and intraobserver agreement were substantial (kappa=0.69 and 0.78, respectively). CONCLUSIONS: Ventricular pointing is a highly agreed-on finding that is substantially associated with but not pathognomonic for fetal spina bifida.

The effects of misinterpretation of an artefact on multidetector row CT scans in children.

BACKGROUND: Artefacts reflect problems with radiographic technique rather than true pathology. These may be misinterpreted as pathology with serious consequences. An artefact caused such problems in one paediatric imaging department. OBJECTIVE: To determine the incidence, and consequences of misinterpretation, of a CT artefact in a paediatric imaging department. MATERIALS AND METHODS: A retrospective review of images and reports of paediatric CT scans over a set period with a known artefact was performed. Reports were correlated with reviewers' evaluation of the presence of artefact and reviewed for correct identification of artefact, misinterpretation as pathology, and action taken as a result. RESULTS: A total of 74 CT scans had been performed over the study period and an artefact detected by reviewers on 32 (43%). Six (18.75%) of these were misinterpreted as pathology, of which three (9.4%) were reported as tuberculous granulomas, two (6.2%) as haemorrhages and one (3.1%) as an unknown hyperdensity. Two patients (6.2%) had subsequent MRI studies performed, and treatment for tuberculosis was continued in one patient (3.1%). CONCLUSION: No initial report identified the artefact. One-fifth of the scans with the artefact were misinterpreted as pathology and half of these misinterpretations led to further action. Artefacts result in false diagnoses and unnecessary investigations; vigilance is needed.

Prevalence and evolution of intracranial hemorrhage in asymptomatic term infants.

BACKGROUND AND PURPOSE: Subdural hemorrhage (SDH) is often associated with infants experiencing nonaccidental injury (NAI). A study of the appearance and natural evolution of these birth-related hemorrhages, particularly SDH, is important in the forensic evaluation of NAI. The purpose of this study was to determine the normal incidence, size, distribution, and natural history of SDH in asymptomatic term neonates as detected by sonography (US) and MR imaging within 72 hours of birth. MATERIALS AND METHODS: Birth history, delivery method, duration of each stage of labor, pharmaceutic augmentation, and complications during delivery as well as postnatal physical examination were recorded. Brain MR imaging and US were performed on 101 asymptomatic term infants at 3-7 days, 2 weeks, 1 month, and 3 months. Clinical follow-up at 24 months was recorded. RESULTS: Forty-six neonates had SDH by MR imaging within 72 hours of delivery. SDH was seen in both vaginal and cesarean deliveries. All neonates were asymptomatic, with normal findings on physical examination. All 46 had supratentorial SDH seen in the posterior cranium. Twenty (43%) also had infratentorial SDH. US detected 11 of the 20 (55%) infratentorial SDHs and no supratentorial SDH. Most SDHs present at birth were