James M Drake

SickKids, Toronto, Ontario, Canada

Are you James M Drake?

Claim your profile

Publications (321)756.27 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: CNS germinomas have an excellent prognosis with radiation therapy alone. However, in children, volume and dose of CNS radiation are associated with neurocognitive and neuroendocrine sequelae. Our objective was to determine long-term outcomes of our cohort who received chemotherapy and reduced radiation. This retrospective cohort study analyzed treatment and outcome of intracranial germinoma patients consecutively treated at Sick Kids, Toronto, Canada, from January 2000 to December 2013. 24 children (13 male, 11 female; median age 13.36 years) were identified. Median follow up was 61 months (range 1-144 months). Tumor location was suprasellar (n = 9), bifocal (8), pineal (6), and basal ganglia (1). Three children showed dissemination on imaging. 2/24 had only elevated serum human chorionic gonadotropin, 3/24 only elevated lumbar cerebrospinal fluid (CSF) hCG, and 2/24 had both elevated serum and lumbar CSF hCG. 23/24 children completed treatment and received multi-agent chemotherapy followed by either ventricular radiation (2340-2400 cGy) (n = 9), ventricular radiation + boost (1600 cGy) (n = 8), whole brain (2340 cGy) (n = 3), focal (4000 cGy) (n = 2) or craniospinal radiation (2340 cGy) (n = 1). Five-year progression free and overall survival was 96 and 100 % respectively. 8/24 patients with ventricular radiation ± boost (2340/4000 cGy) displayed stable full scale intelligence quotient over a mean interval of 3 years following radiation, but showed declined processing speed. In this limited experience, excellent 5-year overall survival rates were achieved with chemotherapy followed by reduced whole ventricular radiation even if ventricular radiation was delivered without boost.
    No preview · Article · Jan 2016 · Journal of Neuro-Oncology
  • Charles Mougenot · Adam Waspe · Thomas Looi · James M Drake
    [Show abstract] [Hide abstract]
    ABSTRACT: Magnetic resonance acoustic radiation force imaging (MR-ARFI) allows the quantification of microscopic displacements induced by ultrasound pulses, which are proportional to the local acoustic intensity. This study describes a new method to acquire MR-ARFI maps, which reduces the measurement noise in the quantification of displacement as well as improving its robustness in the presence of motion. Two MR-ARFI sequences were compared in this study. The first sequence 'variable MSG' involves switching the polarity of the motion sensitive gradient (MSG) between odd and even image frames. The second sequence named 'static MSG' involves a variable ultrasound trigger delay to sonicate during the first or second MSG for odd and even image frames, respectively. As previously published, the data acquired with a variable MSG required the use of reference data acquired prior to any sonication to process displacement maps. In contrary, data acquired with a static MSG were converted to displacement maps without using reference data acquired prior to the sonication. Displacement maps acquired with both sequences were compared by performing sonications for three different conditions: in a polyacrylamide phantom, in the leg muscle of a freely breathing pig and in the leg muscle of pig under apnea. The comparison of images acquired at even image frames and odd image frames indicates that the sequence with a static MSG provides a significantly better steady state (p < 0.001 based on a Student's t-test) than the images acquired with a variable MSG. In addition no reference data prior to sonication were required to process displacement maps for data acquired with a static MSG. The absence of reference data prior to sonication provided a 41% reduction of the spatial distribution of noise (p < 0.001 based on a Student's t-test) and reduced the sensitivity to motion for displacements acquired with a static MSG. No significant differences were expected and observed for thermal maps acquired with a variable MSG and a static MSG. The use of a static MSG with a variable ultrasound trigger delay improves the ARFI displacement map quality without additional acquisition time and remains compatible with the simultaneous acquisition of MR thermal maps.
    No preview · Article · Dec 2015 · Physics in Medicine and Biology
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: OBJECT In a previous report by the same research group (Kestle et al., 2011), compliance with an 11-step protocol was shown to reduce CSF shunt infection at Hydrocephalus Clinical Research Network (HCRN) centers (from 8.7% to 5.7%). Antibiotic-impregnated catheters (AICs) were not part of the protocol but were used off protocol by some surgeons. The authors therefore began using a new protocol that included AICs in an effort to reduce the infection rate further. METHODS The new protocol was implemented at HCRN centers on January 1, 2012, for all shunt procedures (excluding external ventricular drains [EVDs], ventricular reservoirs, and subgaleal shunts). Procedures performed up to September 30, 2013, were included (21 months). Compliance with the protocol and outcome events up to March 30, 2014, were recorded. The definition of infection was unchanged from the authors' previous report. RESULTS A total of 1935 procedures were performed on 1670 patients at 8 HCRN centers. The overall infection rate was 6.0% (95% CI 5.1%-7.2%). Procedure-specific infection rates varied (insertion 5.0%, revision 5.4%, insertion after EVD 8.3%, and insertion after treatment of infection 12.6%). Full compliance with the protocol occurred in 77% of procedures. The infection rate was 5.0% after compliant procedures and 8.7% after noncompliant procedures (p = 0.005). The infection rate when using this new protocol (6.0%, 95% CI 5.1%-7.2%) was similar to the infection rate observed using the authors' old protocol (5.7%, 95% CI 4.6%-7.0%). CONCLUSIONS CSF shunt procedures performed in compliance with a new infection prevention protocol at HCRN centers had a lower infection rate than noncompliant procedures. Implementation of the new protocol (including AICs) was associated with a 6.0% infection rate, similar to the infection rate of 5.7% from the authors' previously reported protocol. Based on the current data, the role of AICs compared with other infection prevention measures is unclear.
    Full-text · Article · Dec 2015 · Journal of Neurosurgery Pediatrics
  • Source
    Gerben E Breimer · Vivek Bodani · Thomas Looi · James M Drake

    Full-text · Dataset · Dec 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECT Hemispherectomy for unilateral, medically refractory epilepsy is associated with excellent long-term seizure control. However, for patients with recurrent seizures following disconnection, workup and investigation can be challenging, and surgical options may be limited. Few studies have examined the role of repeat hemispherotomy in these patients. The authors hypothesized that residual fiber connections between the hemispheres could be the underlying cause of recurrent epilepsy in these patients. Diffusion tensor imaging (DTI) was used to test this hypothesis, and to target residual connections at reoperation using neuronavigation. METHODS The authors identified 8 patients with recurrent seizures following hemispherectomy who underwent surgery between 1995 and 2012. Prolonged video electroencephalography recordings documented persistent seizures arising from the affected hemisphere. In all patients, DTI demonstrated residual white matter association fibers connecting the hemispheres. A repeat craniotomy and neuronavigation-guided targeted disconnection of these residual fibers was performed. Engel class was used to determine outcome after surgery at a minimum of 2 years of follow-up. RESULTS Two patients underwent initial hemidecortication and 6 had periinsular hemispherotomy as their first procedures at a median age of 9.7 months. Initial pathologies included hemimegalencephaly (n = 4), multilobar cortical dysplasia (n = 3), and Rasmussen's encephalitis (n = 1). The mean duration of seizure freedom for the group after the initial procedure was 32.5 months (range 6-77 months). In all patients, DTI showed limited but definite residual connections between the 2 hemispheres, primarily across the rostrum/genu of the corpus callosum. The median age at reoperation was 6.8 years (range 1.3-14 years). The average time taken for reoperation was 3 hours (range 1.8-4.3 hours), with a mean blood loss of 150 ml (range 50-250 ml). One patient required a blood transfusion. Five patients are seizure free, and the remaining 3 patients are Engel Class II, with a minimum follow-up of 24 months for the group. CONCLUSIONS Repeat hemispherotomy is an option for consideration in patients with recurrent intractable epilepsy following failed surgery for catastrophic epilepsy. In conjunction with other modalities to establish seizure onset zones, advanced MRI and DTI sequences may be of value in identifying patients with residual connectivity between the affected and unaffected hemispheres. Targeted disconnection of these residual areas of connectivity using neuronavigation may result in improved seizure outcomes, with minimal and acceptable morbidity.
    No preview · Article · Dec 2015 · Journal of Neurosurgery Pediatrics
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECT The rate of CSF shunt failure remains unacceptably high. The Hydrocephalus Clinical Research Network (HCRN) conducted a comprehensive prospective observational study of hydrocephalus management, the aim of which was to isolate specific risk factors for shunt failure. METHODS The study followed all first-time shunt insertions in children younger than 19 years at 6 HCRN centers. The HCRN Investigator Committee selected, a priori, 21 variables to be examined, including clinical, radiographic, and shunt design variables. Shunt failure was defined as shunt revision, subsequent endoscopic third ventriculostomy, or shunt infection. Important a priori-defined risk factors as well as those significant in univariate analyses were then tested for independence using multivariate Cox proportional hazard modeling. RESULTS A total of 1036 children underwent initial CSF shunt placement between April 2008 and December 2011. Of these, 344 patients experienced shunt failure, including 265 malfunctions and 79 infections. The mean and median length of follow-up for the entire cohort was 400 days and 264 days, respectively. The Cox model found that age younger than 6 months at first shunt placement (HR 1.6 [95% CI 1.1-2.1]), a cardiac comorbidity (HR 1.4 [95% CI 1.0-2.1]), and endoscopic placement (HR 1.9 [95% CI 1.2-2.9]) were independently associated with reduced shunt survival. The following had no independent associations with shunt survival: etiology, payer, center, valve design, valve programmability, the use of ultrasound or stereotactic guidance, and surgeon experience and volume. CONCLUSIONS This is the largest prospective study reported on children with CSF shunts for hydrocephalus. It confirms that a young age and the use of the endoscope are risk factors for first shunt failure and that valve type has no impact. A new risk factor-an existing cardiac comorbidity-was also associated with shunt failure.
    No preview · Article · Dec 2015 · Journal of Neurosurgery Pediatrics
  • Kyle W. Eastwood · Vivek P. Bodani · James M. Drake
    [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: Recent innovations to expand the scope of intraventricular neuroendoscopy have focused on transitioning multiple-incision procedures into single-corridor approaches. However, the successful adoption of these combined procedures requires minimizing the unwanted torques applied to surrounding healthy structures. OBJECTIVE: To define the geometry of relevant anatomical structures in endoscopic third ventriculostomy (ETV) and pineal region tumor biopsy (ETB). Second, to determine the optimal instrument shaft path required for collision-free single burr hole combined ETV/ETB. METHODS: Magnetic resonance and computed tomography data from 15 pediatric patients who underwent both ETV and ETB procedures between 2006 and 2014 was segmented by using the 3DSlicer software package to create virtual 3-D patient models. Anatomical regions of interest were measured including the foramen of Monro, the massa intermedia, the floor of the third ventricle, and the tumor margin. Utilizing the MATLAB software package, virtual dexterous instruments were inserted into the models and optimal dimensions were calculated. RESULTS: The diameters of the foramen of Monro, massa intermedia (anterior-posterior, superior-inferior), anterior third ventricle, and tumor margin are 6.85, 4.01, 5.05, 14.2, and 28.5 mm, respectively. The average optimal burr placement was determined to be 22.5 mm anterior to the coronal and 30 mm lateral to the sagittal sutures. Optimal flexible instrument geometries for novel instruments were calculated. CONCLUSION: We have established a platform for estimating the shape of novel curved dexterous instruments for collision-free targeting of multiple intraventricular points, which is both patient and tool specific and can be integrated with image guidance. These data will aid in developing novel dexterous instruments. ABBREVIATIONS: DICOM, Digital Imaging and Communications in Medicine ETB, endoscopic tumor biopsy ETV, endoscopic third ventriculostomy FM, foramen of Monro KMKT, Knaus, Matthias, Koch, and Thomale MI, massa intermedia STL, Standard Template Library
    No preview · Article · Dec 2015 · Neurosurgery
  • Source

    Full-text · Article · Nov 2015 · Neurosurgery

  • No preview · Conference Paper · Sep 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: Epilepsy surgery can be successful in children with extensive congenital or early acquired focal or hemispheric brain lesion on magnetic resonance imaging (MRI) despite generalized interictal epileptiform discharges (IEDs). The aim of this study was to assess if rapid eye movement (REM) sleep reduced generalized IEDs and revealed lateralized IEDs to identify the epileptogenic hemisphere in children with generalized IEDs and normal/subtle changes on MRI. We studied 20 children with generalized IEDs on scalp electroencephalography (EEG) and normal/subtle changes on MRI who underwent intracranial video-EEG for epilepsy surgery. We assessed a minimum of 100 IEDs during REM, non-REM, and wakefulness, and assigned the distribution (generalized, left, or right hemisphere) to each IED. The number of lobes in the resected areas and seizure outcome were compared between 20 children with generalized IEDs and a comparison group of 28 children without generalized IEDs. The mean occurrence rate of generalized IEDs during REM (37%) was significantly lower than that during non-REM (67%, p < 0.001) and wakefulness (54%, p = 0.003). The number of children whose largest number of IEDs was lateralized in REM was significantly higher than that in non-REM (15 vs. 3 children, 75% vs. 15%, p < 0.001). The hemisphere with lateralized IEDs among three states corresponded with the surgical side in 16 children with generalized IEDs. Seventeen children (85%) with generalized IEDs and 27 (96%) without generalized IEDs underwent resective surgery. Multilobar resection was required for 16 children (94%) with generalized IEDs more frequently than 7 children (26%) without generalized IEDs (p < 0.001). Thirteen children (77%) with generalized IEDs and 19 (73%) without generalized IEDs were seizure-free with a mean of 3.3 years of follow-up. Our study demonstrates the importance of assessing REM in children with generalized IEDs as it reveals lateralized epileptogenic spikes. Seizure freedom may be achieved with multilobar resection in these children with generalized IEDs and normal/subtle changes on MRI. Generalized IEDs in children with normal/subtle changes on MRI should not preclude surgical resection. Wiley Periodicals, Inc. © 2015 International League Against Epilepsy.
    No preview · Article · Jul 2015 · Epilepsia
  • [Show abstract] [Hide abstract]
    ABSTRACT: Cranio-orbital remodeling aims to correct the dysmorphic skull associated with craniosynostosis. Traditionally, the skull is reconstructed into a shape that is subjectively normal according to the surgeon's perception. We present a novel technique using a mathematical algorithm to define the optimal location for bony osteotomies and to objectively reshape the fronto-orbital bar into an ideal normal skull contour.Using pre-operative computed tomography images, the abnormal skull contour at the frontal-orbital region was obtained for infants planned to undergo cranio-orbital remodeling. The ideal skull shape was derived from an age- and sex-matched normative skull library. For each patient, the mathematical technique of dynamic programming (DP) was applied to compare the abnormal and ideal skull shapes. The DP algorithm identifies the optimal location of osteotomy sites and calculates the objective difference in surface area remaining between the normative and dysmorphic skull shape for each solution applied. By selecting the optimal solution with minimal objective difference, the surgeon is guided to reproducibly recreate the normal skull contour with defined osteotomies.The DP algorithm was applied in 13 cases of cranio-orbital remodeling. Five female and 8 male infants with a mean age of 11 months were treated for craniosynostosis classified as metopic (n = 7), unicoronal (n = 4), or bicoronal (n = 2). The mean OR time was 190.2 min (SD 33.6), mean estimated blood loss 244 cc (SD 147.6), and 10 infants required blood transfusions. Compared with a historical crania-orbital remodeling group treated without application of the algorithm, there was no significant difference in OR time, estimated blood loss, or transfusion rate.This novel technique enables the craniofacial surgeon to objectively reshape the fronto-orbital bar and reproducibly reconstruct a skull shape resembling that of normal infants.
    No preview · Article · Jul 2015 · The Journal of craniofacial surgery
  • Source

    Preview · Article · Jun 2015
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This study aims to develop and establish the content validity of multiple expert rating instruments to assess performance in endoscopic third ventriculostomy (ETV), collectively called the Neuro-Endoscopic Ventriculostomy Assessment Tool (NEVAT). The important aspects of ETV were identified through a review of current literature, ETV videos, and discussion with neurosurgeons, fellows, and residents. Three assessment measures were subsequently developed: a procedure-specific checklist (CL), a CL of surgical errors, and a global rating scale (GRS). Neurosurgeons from various countries, all identified as experts in ETV, were then invited to participate in a modified Delphi survey to establish the content validity of these instruments. In each Delphi round, experts rated their agreement including each procedural step, error, and GRS item in the respective instruments on a 5-point Likert scale. Seventeen experts agreed to participate in the study and completed all Delphi rounds. After item generation, a total of 27 procedural CL items, 26 error CL items, and 9 GRS items were posed to Delphi panelists for rating. An additional 17 procedural CL items, 12 error CL items, and 1 GRS item were added by panelists. After three rounds, strong consensus (>80 % agreement) was achieved on 35 procedural CL items, 29 error CL items, and 10 GRS items. Moderate consensus (50-80 % agreement) was achieved on an additional 7 procedural CL items and 1 error CL item. The final procedural and error checklist contained 42 and 30 items, respectively (divided into setup, exposure, navigation, ventriculostomy, and closure). The final GRS contained 10 items. We have established the content validity of three ETV assessment measures by iterative consensus of an international expert panel. Each measure provides unique assessment information and thus can be used individually or in combination, depending on the characteristics of the learner and the purpose of the assessment. These instruments must now be evaluated in both the simulated and operative settings, to determine their construct validity and reliability. Ultimately, the measures contained in the NEVAT may prove suitable for formative assessment during ETV training and potentially as summative assessment measures during certification.
    Full-text · Article · May 2015 · Child s Nervous System
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: Clinical trials for Transcranial MR-guided Focused Ultrasound (TcMRgFUS) treatments are ongoing for brain tumor and essential tremor therapies in adult patients. However, applications for children have not been proposed despite their thinner skull being potentially less of an acoustic barrier and the presence of an open fontanelle on neonates. TcMRgFUS systems designed for adults are not particularly suited for treatments in the youngest patients,but the favorable acoustic properties of their skulls could allow the use of less complex systems. As there is minimal data in the literature on the evolution of the acoustic properties of the skull with growth, the aim of this study was to perform the acoustic characterization of pediatric skulls using a clinical MR-guided HIFU system to determine the feasibility of treatments according to patient age. METHODS: This study was performed on a neonate and an 8-year-old (8-y.) cadaveric skull. A 0.2 mm needle hydrophone was placed in a tank of degassed water and aligned to the geometric focus of a clinical HIFU transducer (Philips V1 Sonalleve). The signals of the 256 elements of the phased array transducer were acquired as a baseline measurement using the hydrophone and the MatHIFU software toolkit. The degassed skull was placed inside the tank between the hydrophone and the transducer (Figure 1). New acquisitions were performed for different angular orientations of the skull, according to the sagittal and coronal axes, in the range of ± 15°, allowing access to different areas of interest. Maps of insertion losses (IL) and time-of-flight (TOF) delays due to the skull were deducted from these measurements performed at 1 MHz, 1.2 MHz and 1.45 MHz. RESULTS: The fontanelle, the parietal bones and the frontal bones were characterised on the neonate skull. The average value (± s.d.) of the TOF delay for the fontanelle was -0.08 ± 0.02 µs over the frequency range. The average value (± s.d.) of the IL was 1.1 ± 1.0, 0.5 ± 0.6 and 0.9 ± 0.8 dB for the frequencies of 1, 1.2 and 1.45 MHz respectively. For the parietal bones, in the same order of frequencies, the TOF delays were -0.13 ± 0.02, -0.16 ± 0.01 and -0.16 ± 0.01 µs respectively and the IL were 2.8 ± 1.0, 1.2 ± 0.5 and 1.6 ± 0.7 dB respectively. For the frontal bones, in the same order of frequencies, the TOF delays were –0.14 ± 0.02, -0.18 ± 0.03 and -0.17 ± 0.02 µs respectively and the IL were 3.2 ± 1.2, 2.0 ± 0.7 and 2.9 ± 1.1 dB respectively. On the 8-y. skull, the sagittal suture and the left and right parietal bones were studied. For the sagittal suture, at 1.2 MHz, the TOF delay was -0.78 ± 0.22 µs and the IL was 14.2 ± 2.1 dB. For the right and left parietal bones, at 1.2 MHz, the TOF delays were -0.79 ± 0.11 and -0.71 ± 0.06 µs respectively and the IL were 11.5 ± 1.6 and 8.3 ± 1.2 dB respectively. Measurements at 1 and 1.45 MHz for this skull are ongoing. CONCLUSIONS: On the neonate skull, the fontanelle can really be described as an acoustic window. On the 8-y. skull, the sagittal suture leads to higher IL than the rest of the skull but does not affect the TOF delay. As expected, the 8-y. skull bones have higher IL and TOF delays than the neonate ones but are still low compared to adults data found in the literature. Detailed mappings of the acoustic properties of two pediatric skulls were obtained using setting close to clinical conditions. These results confirm the potential of this technology for neurological disorder treatments in pediatric patients. Funding provided by Brain Canada, and the Focused Ultrasound Foundation. In-kind technical support provided by Philips Healthcare.
    No preview · Conference Paper · Apr 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: Introduction: Magnetic resonance guided high intensity focused ultrasound (MRgHIFU) is a non-invasive alternative to surgical intervention and could be used to treat fetal and placental conditions in utero. Potential circumstances involve abnormal extra tissue burden or irregular blood supply that occurs in twin-to-twin transfusion syndrome or placental and fetal benign and cystic tissue disorders. This pilot study assesses the feasibility of in utero MRgHIFU delivery in a late-gestation rabbit model and determines the safety, accuracy and tissue response to the treatment. Methods: Under preanasthetic, a late gestation rabbit (~30 days gestation, term = 32; n=2) was intubated and put under ~3% isoflurane with 2 L of oxygen. The abdominal area was shaved and depilated, covered with degassed ultrasound gel and the rabbit carefully placed in a prone position coupling the abdomen to a gel pad on the Phillips V1 Sonalleve HIFU table. The rabbit received pancuronium bromide (1ml/kg) to reduce movement during MR imaging and HIFU sonication. High resolution 3D balanced field echo anatomical images (bFFE, voxel size = 0.8mm x 0.8mm x 1mm, TE = 2.3ms, TR = 4.6ms, FA = 35°, NSA = 2), acquired on a Phillips Achieva 3T MR system were used to identify fetuses in the treatment envelope of the HIFU transducer. These images were also used to select specific placental and fetal targets including the lung, liver, sacrum, umbilical cord, brain and heart targets [figure 1A]. Clusters of HIFU sonications were applied continuously for ~20 seconds at a transmission frequency of 1.2 MHz at various powers to these targets. Standard MR thermometry measured using the proton resonance frequency shift of water, produced accurate (<1°C) and dynamic (<1s) thermal maps in soft tissue in response to HIFU sonication. All HIFU exposures are presented as average acoustic power (watts) ± standard deviation reaching maximum temperatures in degrees Celsius of the treatment. After post-treatment imaging (bFFE), treatment and control fetuses (n=10) underwent full necropsy and tissues were processed, paraffin embedded and stained with hematoxylin and eosin for histological analysis. Results: Placentas with localized HIFU cells (n=4) were targeted at 35±15W reaching 64°C, 62.5±13.9W reaching 70°C, 68.3±18.6W reaching 75°C and 65±35W reaching 49° respectively. The histological changes in these placentas were circumscribed areas of hemorrhage and necrotic tissue [figure 1B]. Placentas with bisecting linear HIFU cells (n=2) were targeted at 92.3±50W reaching 59° and 133.8±43W reaching 69°. The attempted bisection resulted in diffuse necrosis in maternal and fetal placenta and hemorrhage throughout. The brain was exposed to 40±10W reaching 74°C in the thalamus, the umbilical cord was exposed to 35±15W reaching 50° and the sacrum was exposed to 30W reaching 78°, and resulted in no visible histological changes. The lungs (n=2) were exposed to 70W reaching 75°C and 100W reaching 90° which caused hemorrhagic foci with visible necrotic areas showing damage to large vessels, bronchi and alveoli. Lastly, the heart was exposed to 70W reaching 74°C resulting in vascular engorgement, focal disruption and hemorrhage into the myocardium with wavy fibers and evidence of early necrosis. Conclusions: The late gestation rabbit model suitably delineates targets and evaluates MRgHIFU applications in utero as treatment of fetal lesions and abnormal placental circulation. Future studies will focus on further characterizing and quantifying HIFU induced tissue changes and exploring feasibility of survival experiments to determine fetus viability post treatment.
    No preview · Conference Paper · Apr 2015
  • T. Looi · A. Waspe · C. Mougenot · J. Amaral · K. Hynynen · J. Drake
    [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVES: To determine the efficacy of MRgFUS thrombolysis for the treatment of IVH clots and subsequent post hemorrhagic ventricular dilation in acute and chronic porcine model as an analogue for IVH in prematurity. METHODS: Following baseline imaging (T1­, T2 and T2* MR at 3T (Phillips Acheiva)), IVH was created by injecting 1.25cc/kg of fresh autologous blood harvested from a femoral vessel into the right lateral ventricle of the brain of five neonatal piglets ~3 kg in weight. The entry site was 1cm lateral, and 1 cm posterior to the coronal suture. Intra­cranial pressure was measured using a Codman transducer inserted through a similar burr hole on the opposite side. Blood was infused via a 16G angiocatheter at a rate of 0.3cc/min. Repeat imaging sequences verified clot location and size. Animal was recovered and a neurological score was assessed to monitor the behavior. For the thrombolytic treatment protocol 7 post days injection, a 4­5 cm2 craniotomy to simulate the fontanelle was fashioned. A mixture of de­gassed ultrasound gel and water was placed on the brain and the scalp sutured closed. The piglet was oriented supine and feet first on the MRI where the head was placed on Philips Sonalleve FUS transducer, with bilateral Flex­S/M coils. The IVH clots were targeted using 2mm diameter treatment cells at 0.1% to 1% duty cycle and 450 to 900W, pulse duration of 17­133 microsecronds. During treatment, proton resonant frequency (PRF) thermometry and acoustic radiation force imaging (ARFI) was used to monitor temperature change and spatial displacement. Subsequent MR imaging was performed every at 7, 14 and 21 days to monitor changes in the ventricle and re­absorption of clot. RESULTS: For acute studies, 5 piglets were completed where there was an average clot size of 0.718 cc. The intraventricular clot injection was generally well tolerated. Post­infusion ventricular enlargement was substantive (ventricle height increased from 3.2mm to 8.1mm) with noted change in animal behavior at day 10­11. For the thrombolysis treatment, at power level < 600 W and duty cycle < 0.4%, there was no change in clot structure and temperature. At duty cycle > 0.8%, there was a noticeable change in the clot shape, image and an associated temperature increase of ~3­4C with PRF imaging. ARFI revealed that the clot underwent displacements of up to 50 micrometres and confirmed the targeting was correctly located. With the optimized acoustic parameters, there was significant changes in the clot imaging and structure (see Figure). Post mortem extracted clots showed evidence of cavitation with a distinct colour and structural change. CONCLUSIONS: An IVH porcine model was successfully created for both acute and chronic studies. The piglets tolerated the clot creation, and developed hydrocephalus over the 21 day period. MRgFUS sonothrombolysis reduced clot volume with evidence of lysis. Additional work will quantify the evolution of ventricular size post MRgFUS thrombolysis, and quantify acoustic refocusing across the piglet skull.
    No preview · Conference Paper · Apr 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: As interest in applying cognitive load theory (CLT) to the study and design of pedagogic and technological approaches in healthcare simulation grows, suitable measures of cognitive load (CL) are needed. Here, we report a two-phased study investigating the sensitivity of subjective ratings of mental effort (SRME) and secondary-task performance (signal detection rate, SDR and recognition reaction time, RRT) as measures of CL. In phase 1 of the study, novice learners and expert surgeons attempted a visual-monitoring task under two conditions: single-task (monitoring a virtual patient's heart-rate) and dual-task (tying surgical knots on a bench-top simulator while monitoring the virtual patient's heart-rate). Novices demonstrated higher mental effort and inferior secondary-task performance on the dual-task compared to experts (RRT 1.76 vs. 0.73, p = 0.012; SDR 0.27 vs. 0.97, p < 0.001; SRME 7.75 vs. 2.80, p < 0.001). Similarly, secondary task performance deteriorated from baseline to dual-task among novices (RRT 0.63 vs. 1.76 s, p < 0.006 and SDR 1.00 vs. 0.27, p < 0.001), but not experts (RRT 0.63 vs. 0.73 s, p = 0.124 and SDR 1.00 vs. 0.97, p = 0.178). In phase 2, novices practiced surgical knot-tying on the bench top simulator during consecutive dual-task trials. A significant increase in SDR (F(9,63) = 6.63, p < 0.001, f = 0.97) and decrease in SRME (F(9,63) = 9.39, p < 0.001, f = 1.04) was observed during simulation training, while RRT did not change significantly (F(9,63) = 1.18, p < 0.32, f = 0.41). The results suggest subjective ratings and dual-task performance can be used to track changes in CL among novices, particularly in early phases of simulation-based skills training. The implications for measuring CL in simulation instructional design research are discussed.
    No preview · Article · Mar 2015 · Advances in Health Sciences Education
  • M. Hess · T. Looi · A. Lasso · G. Fichtinger · J. Drake
    [Show abstract] [Hide abstract]
    ABSTRACT: Purpose: Intraventricular hemorrhage (IVH) affects nearly 15% of preterm infants. It can lead to ventricular dilation and cognitive impairment. To ablate IVH clots, MR-guided focused ultrasound surgery (MRgFUS) is investigated. This procedure requires accurate, fast and consistent quantification of ventricle and clot volumes. Methods: We developed a semi-autonomous segmentation (SAS) algorithm for measuring changes in the ventricle and clot volumes. Images are normalized, and then ventricle and clot masks are registered to the images. Voxels of the registered masks and voxels obtained by thresholding the normalized images are used as seed points for competitive region growing, which provides the final segmentation. The user selects the areas of interest for correspondence after thresholding and these selections are the final seeds for region growing. SAS was evaluated on an IVH porcine model. Results: SAS was compared to ground truth manual segmentation (MS) for accuracy, efficiency, and consistency. Accuracy was determined by comparing clot and ventricle volumes produced by SAS and MS, and comparing contours by calculating 95% Hausdorff distances between the two labels. In Two-One-Sided Test, SAS and MS were found to be significantly equivalent (p < 0.01). SAS on average was found to be 15 times faster than MS (p < 0.01). Consistency was determined by repeated segmentation of the same image by both SAS and manual methods, SAS being significantly more consistent than MS (p < 0.05). Conclusion: SAS is a viable method to quantify the IVH clot and the lateral brain ventricles and it is serving in a largescale porcine study of MRgFUS treatment of IVH clot lysis.
    No preview · Article · Jan 2015
  • [Show abstract] [Hide abstract]
    ABSTRACT: We analyzed the spatial distribution and concordance of fast (>10Hz) and slow (<5Hz) electroencephalogram (EEG) components of ictal activities and interictal epileptiform discharges (IIED) recorded by intracranial video EEG (IVEEG) in children with epileptic spasms (ES). We studied eight children with ES, who underwent IVEEG before resective surgery for epilepsy. We quantified the root-mean-square (RMS) amplitude of the fast and slow components of ictal activities during ES and IIED. We compared the concordance between the spatial distributions of the fast and slow components of ES and IIED. There was a larger concordance between the spatial distributions of the fast and slow components in IIED than in ES (p=0.0206 and 0.0401). The spatial concordance between the fast and slow EEG components was significantly different between ES and IIED. The mechanisms underlying the generation of slow EEG components may differ between ES and IIED. The slow EEG components of ES might indicate an extensive epileptic network involving remote symptomatic zones for ES in either the cortical or subcortical areas. The high spatial concordance between the fast and slow components of IIED suggests the involvement of a local inhibitory process within the epileptic cortex. Copyright © 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
    No preview · Article · Dec 2014 · Clinical neurophysiology: official journal of the International Federation of Clinical Neurophysiology
  • H. Azimian · T. Looi · J. Drake
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, a novel closed-loop inverse kinematics scheme that is capable of handling inequality constraints is proposed. By a proper adoption of slack variables and using the task priority concept, the proposed method allows for the accommodation of inequality constraints in a closed-form inverse kinematics solution for robotic manipulators. Unlike other formulations, the proposed scheme does not require solving a sequence of constrained optimization problems in real time and in this sense is faster. The performance of the proposed scheme is investigated through an implementation for real-time control of a concentric-tube robot with a restricted operational space. The experimental results demonstrate the efficacy of the scheme in handling inequality constraints in real-time closed-loop inverse kinematics.
    No preview · Article · Oct 2014

Publication Stats

8k Citations
756.27 Total Impact Points


  • 1993-2015
    • SickKids
      • Division of Neurosurgery
      Toronto, Ontario, Canada
    • Georgetown University
      • Department of Plastic Surgery
      Washington, Washington, D.C., United States
  • 1991-2015
    • University of Toronto
      • • Division of Neurosurgery
      • • Department of Surgery
      • • Department of Psychology
      • • Hospital for Sick Children
      • • Institute of Biomaterials and Biomedical Engineering
      Toronto, Ontario, Canada
  • 2011-2012
    • Toronto Western Hospital
      Toronto, Ontario, Canada
  • 2008
    • University of North Carolina at Chapel Hill
      • Department of Neurosurgery
      Chapel Hill, NC, United States
  • 2003
    • Dalhousie University
      • Division of Neurosurgery
      Halifax, Nova Scotia, Canada
  • 2001
    • Concordia University–Ann Arbor
      Ann Arbor, Michigan, United States
  • 1994
    • University of Waterloo
      • Department of Applied Mathematics
      Ватерлоо, Ontario, Canada