Imaging plays a fundamental role in the managing childhood neurologic, neurosurgical and neuro-oncological disease. Employing multi-parametric MRI techniques, such as spectroscopy and diffusion- and perfusion-weighted imaging, to the radiophenotyping of neuroradiologic conditions is becoming increasingly prevalent, particularly with radiogenomic analyses correlating imaging characteristics with molecular biomarkers of disease. However, integration into routine clinical practice remains elusive. With modern multi-parametric MRI now providing additional data beyond anatomy, informing on histology, biology and physiology, such metric-rich information can present as information overload to the treating radiologist and, as such, information relevant to an individual case can become lost. Artificial intelligence techniques are capable of modelling the vast radiologic, biological and clinical datasets that accompany childhood neurologic disease, such that this information can become incorporated in upfront prognostic modelling systems, with artificial intelligence techniques providing a plausible approach to this solution. This review examines machine learning approaches than can be used to underpin such artificial intelligence applications, with exemplars for each machine learning approach from the world literature. Then, within the specific use case of paediatric neuro-oncology, we examine the potential future contribution for such artificial intelligence machine learning techniques to offer solutions for patient care in the form of decision support systems, potentially enabling personalised medicine within this domain of paediatric radiologic practice.
19 September 2016 in British Journal Of Neurosurgery
Intracranial pressure (ICP) measurement is an important diagnostic tool in Neurosurgery. Until relatively recently, conventional monitoring has required that subjects be admitted to a hospital bed and the device is only able to be left in-situ for limited periods of time. We have evaluated a Telemetric ICP monitoring system that has been proven, by several other groups worldwide, to permit rapid, repeated and prolonged ICP measurement, in multiple environments. In our unit, 4 patients have been implanted to-date, between the ages of 4 and 16, manifesting a wide range of complex neurosurgical conditions. The sensors have been left in-situ for between 460 and 632 days. There have been no clinical complications and the system has been universally well tolerated. Clinical events, costs and patient experience were all assessed prior to and following implantation. Overall, there was a significant reduction in associated admissions (44.3%), imaging requirements (72.5%) and costs (50.0%). Subjective feedback from both the patients (where possible) and their families was overwhelmingly positive, partly due to (a) the system's ease of use, (b) its ability to reduce the number of admissions/tests required and (c) the facility for rapid measurement of ICP that permitted on-the-spot reassurance of concerns. Additionally, the ability to monitor ICP at home and/or whilst ambulant, has provided measurements that were hitherto inaccessible to our team, facilitating all the potential benefits that analysis of such information would provide. Indeed, we have seen the resultant management in each case has been completely altered by the availability of this data, reaffirming that the importance of being able to obtain it should not be underestimated. The combination of both this and the ability to markedly improve patient experience, along with generating significant cost-savings, lead the authors to suggest that the implantation of this system should be strongly considered in selected individuals.
Chimeric mouse models generated via adoptive bone marrow transfer are the foundation for immune cell tracking in neuroinflammation. Chimeras that exhibit low chimerism levels, blood-brain barrier disruption and pro-inflammatory effects prior to the progression of the pathological phenotype, make it difficult to distinguish the role of immune cells in neuroinflammatory conditions. Head-shielded irradiation overcomes many of the issues described and replaces the recipient bone marrow system with donor haematopoietic cells expressing a reporter gene or different pan-leukocyte antigen, whilst leaving the blood-brain barrier intact. However, our previous work with full body irradiation suggests that this may generate a pro-inflammatory peripheral environment which could impact on the brain’s immune microenvironment. Our aim was to compare non-myeloablative busulfan conditioning against head-shielded irradiation bone marrow chimeras prior to implantation of glioblastoma, a malignant brain tumour with a pro-inflammatory phenotype. Recipient wild-type/CD45.1 mice received non-myeloablative busulfan conditioning (25 mg/kg), full intensity head-shielded irradiation, full intensity busulfan conditioning (125 mg/kg) prior to transplant with whole bone marrow from CD45.2 donors and were compared against untransplanted controls. Half the mice from each group were orthotopically implanted with syngeneic GL-261 glioblastoma cells. We assessed peripheral blood, bone marrow and spleen chimerism, multi-organ pro-inflammatory cytokine profiles at 12 weeks and brain chimerism and immune cell infiltration by whole brain flow cytometry before and after implantation of glioblastoma at 12 and 14 weeks respectively. Both non-myeloablative conditioning and head-shielded irradiation achieve equivalent blood and spleen chimerism of approximately 80%, although bone marrow engraftment is higher in the head-shielded irradiation group and highest in the fully conditioned group. Head-shielded irradiation stimulated pro-inflammatory cytokines in the blood and spleen but not in the brain, suggesting a systemic response to irradiation, whilst non-myeloablative conditioning showed no cytokine elevation. Non-myeloablative conditioning achieved higher donor chimerism in the brain after glioblastoma implantation than head-shielded irradiation with an altered immune cell profile. Our data suggest that non-myeloablative conditioning generates a more homeostatic peripheral inflammatory environment than head-shielded irradiation to allow a more consistent evaluation of immune cells in glioblastoma and can be used to investigate the roles of peripheral immune cells and bone marrow-derived subsets in other neurological diseases.
1 December 2015 in Journal of neurosurgery. Pediatrics
OBJECT Endoscopic third ventriculostomy (ETV) has become a widely used method for CSF diversion when treating obstructive hydrocephalus. There are multiple recommendations on the transcortical ETV entry points, and some are specifically designed to provide a trajectory that avoids displacement to the eloquent periventricular structures. However, the morphology of the ventricular system is highly variable in hydrocephalus, and therefore a single best ETV trajectory may not be applicable to all cases. In the current study, 3 frequently quoted ETV entry points are compared in a cohort of pediatric cases with different degrees of ventriculomegaly. METHODS The images of 30 consecutive pediatric patients with varying degrees of ventriculomegaly were reviewed. Three-dimensional models were created using radiological analysis of anatomical detail and preoperative MRI scans in order to simulate 3 frequently quoted ETV trajectories for rigid neuroendoscopes. These trajectories were characterized based on the freque...
Purpose: This study presents a series of paediatric meningiomas and discusses aetiology, risk factors and outcomes with comparison to current literature. Methods: This is a retrospective review of surgically treated paediatric meningiomas from three UK centres: the University Hospital of Wales, Alder Hey Children’s Hospital and Royal Manchester Children’s Hospital. Twenty-seven patients aged 16 and under at the time of their first procedure were identified over a 15-year period (1 January 2007 and 1 March 2023). Electronic medical records were used to collect data on age at presentation, sex, location of tumour(s), extent of resection, histology, WHO grade, complications, outcomes and associated conditions, notably neurofibromatosis type 2 (NF2). Results: Twenty-seven patients underwent 39 procedures. There were 13 males and 14 females. The median age was 13 years (range, 8 months to 16 years). Twenty-one (75%) were WHO grade 1, 6 (21%) were grade 2 and 1 (4%) was grade 3. Eight patients (30%) had confirmed NF2. Twelve patients (44%) were sporadic cases. Twenty-five percent and 50% were the recurrence rate in WHO grade 1 and 2 tumours, respectively. Conclusion: The risk of grade 1 tumour recurrence was higher than within the adult population. This may be due to histological features of paediatric meningiomas differing from the adult population, and therefore, the WHO grading system may not be reflective of recurrence risk. Future molecular profiling and larger studies are required given the rarity of these cases.
Paediatric brain tumours and their treatments are associated with long-term cognitive impairment. While the aetiology of cognitive impairment is complex and multifactorial, multiparametric Magnetic Resonance Imaging (MRI) can identify many risk factors including tumour location, damage to eloquent structures and tumour phenotype. Hydrocephalus and raised intracranial pressure can be observed, along with risk factors for post-operative paediatric cerebellar mutism syndrome or epilepsy. MRI can also identify complications of surgery or radiotherapy and monitor treatment response. Advanced imaging sequences provide valuable information about tumour and brain physiology, but clinical use is limited by extended scanning times and difficulties in processing and analysis. Brain eloquence classifications exist, but focus on adults with neurological deficits and are outdated. For the analysis of childhood tumours, limited numbers within tumour subgroups and the investigation of long-term outcomes necessitate using historical scans and/or multi-site collaboration. Variable imaging quality and differing acquisition parameters limit the use of segmentation algorithms and radiomic analysis. Harmonisation can standardise imaging in collaborative research, but can be challenging, while data-sharing produces further logistical challenges. Consequently, most research consists of small single-centre studies limited to regional analyses of tumour location. Technological advances reducing scanning times increase the feasibility of clinical acquisition of high-resolution standardised imaging including advanced physiological sequences. The RAPNO and SIOPE paediatric brain tumour imaging guidelines have improved image standardisation, which will benefit future collaborative imaging research. Modern machine learning techniques provide more nuanced approaches for integration and analysis of the complex and multifactorial data involved in cognitive outcome prediction.