The fetal period is a time of unparalleled brain growth and development, and is arguably the most important time for defining future cognitive potential. While fetal MRI provides more detailed structural and physiological information compared to ultrasound, fetal motion severely limits the diagnostic capabilities of fetal neuroimaging, and hampers our ability to accurately diagnose disease, select candidates for intervention, and assess response to treatments. In this project, we propose to transform fetal neuroimaging by combining the latest advances in imaging hardware and software to develop a motion insensitive advanced fetal protocol that includes high-resolution anatomical and physiological imaging. Our goal is to transform the field of fetal MRI by developing and employing state-of-the-art advances on the acquisition end of the fetal MRI experiment to meet the growing demand for more information as fetal interventions emerge.

The ChRIS platform is an opensource, distributed platform that standardizes the deployment of container-based computation in arbitrarily complex processing trees. ChRIS strives to accelerate computational innovation by handling most of the complexity of distributed and cloud-based computation, while allowing developers and research to focus on their specific computational problem.  Though geared mostly to medical image data, ChRIS is really data agnostic and can be used to distribute and process any data type and applied to most scientific problems.

The core purpose of this project is to develop a solution for rapidly deploying machine learning pipelines. The emphasis is more on building a containerized end-to-end AI workflow than developing new AI algorithms. Underpinning the deployment of the research is the ChRIS platform which is being used as the mechanism to allows for data set selection, feature extraction, model training and rapid deployment, coupled with continuous learning.

Sickle Cell Disease (SCD) impairs oxygen transport to tissue and causes endothelial injury. Thus, therapeutic interventions aim to improve both, but there is an unmet need for biomarkers to determine when intervention is necessary and evaluate the effectiveness of the chosen intervention in individual patients. We aim to monitor SCD and its treatment through their impact on cerebral hemodynamics, as the brain is one of the most vulnerable and consequential targets of the disease. Specifically, we use quantitative MRI and advanced optical spectroscopy techniques (frequency-domain near-infrared and diffuse correlation spectroscopies; FDNIRS-DCS) to monitor 1) cerebral oxygen transport with measures of cerebral blood flow (CBF), cerebral oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen consumption (CMRO2) and 2) endothelial function with cerebrovascular reactivity (CVR).

The incidence rate of infants developing post-infectious hydrocephalus in sub-Saharan Africa has reached over 100,000 cases per year. In order to address the high need for hydrocephalus treatment in under-resourced countries, Dr. Benjamin Warf has established a training program at CURE Children’s Hospital of Uganda, in which neurosurgeons are trained to perform an endoscopic third ventriculostomy combined with choroid plexus cauterization (ETV/CPC).

The ETV/CPC is the preferred method of hydrocephalus treatment in comparison to shunt methods, as the ETV/CPC is considered lower risk and more cost-effective. While the long-term success rate of ETV/CPC treatment is approximately 70%, around 30% of infants will require further procedures or shunt placement. Thus, a portable, inexpensive, safe, and easy-to-use tool to evaluate the status of hydrocephalus as well as predict treatment outcomes would be of great clinical benefit.

Our sophisticated near-infrared spectroscopy (NIRS) technology measures cerebral blood flow (CBF) and cerebral oxygen metabolism (CMRO2), biomarkers we hypothesize to be indicative of hydrocephalus severity and response to treatment. In conducting this pilot study, we will test the efficacy of NIRS with the goal of establishing it as a bedside diagnostic tool that can be implemented in both developed and developing areas around the world.

Germinal matrix-intraventricular hemorrhage (GM-IVH) occurs in 45% of extremely low birth weight (ELBW) premature infants, often leading to long-term neurodevelopmental impairments (NDI). Post-hemorrhagic hydrocephalus (PHH) is a common complication of GM-IVH and increases the risk of major NDI to 75-90%.

Currently, the only bedside tool available to assess for hemorrhage and secondary hydrocephalus is ultrasound. However, our group has developed advanced near-infrared spectroscopy (NIRS) technology for monitoring cerebral blood flow (CBF) and cerebral oxygen metabolism (CMRO2) in newborns at the bedside. Using FDNIRS-DCS, we have obtained preliminary results showing reduced CBF and CMRO2 values in premature infants with varying grades of GM-IVH.

There exists debate as to whether infants with GM-IVH tend to have typical or worsened neurodevelopmental outcomes when compared to infants without IVH. Our central hypothesis is that NIRS measures of cerebral perfusion and oxygen metabolism are objective criteria for assessing the impact of GM-IVH, the progression of PHH, and the effect of hydrocephalus treatment on CBF and CMRO2. Therefore, we hope to use our technology to settle this controversy by testing if CBF and CMRO2 are indicative of neurodevelopmental outcomes in the IVH and no-IVH populations.

The purpose of this study is to demonstrate the use of Magnetic Resonance Imaging (MRI) in characterizing the effect of nutrition on the development of the infant brain. We aim to be able to identify specific structural and developmental deficiencies and benefits that are associated with different maternal diets. Additionally, we aim to test hypotheses about the effectiveness of infant formulas versus breast feeding in promoting infant brain development, and the mechanisms by which these various nutrient components function in early life.

The purpose of this study is to develop novel MRI acquisition and reconstruction methods, which are faster than the current clinical ones. This new software should improve the quality of the MRI images in terms of how sensitive and specific they are to different tissues, while improving patient comfort. This study consists of one MRI scan, not longer than 1 hour, and is open to children and adults. To find out more information about the study or express your interest, you can email rapidmri@babymri.org or call Sommer Jaber at 617-919-6405.

Tuberous sclerosis complex (TSC) is a rare genetic disorder characterized by growth of benign tumors in the brain and other organs. About half of the TSC patients are also diagnosed with an autism spectrum disorder (ASD). A significant number of individuals diagnosed with TSC, as well as ASD, exhibit language and communication difficulties, but there has been limited research focusing on the language-related brain morphometry in TSC population. Our goal of this retrospective study is to examine the morphological features of different cortical regions that are involved in language processing in patients with TSC, with and without a diagnosis of autism spectrum disorders (ASD).

The purpose of this study is to gain a better understanding on how the potential neurobiological sequelae of chronic childhood adversity shapes the vital adult activity of caring for an infant, and how this in turn affects early brain development. The research team aims to link processes across neurobiological, neuroendocrine, and behavioral levels to examine how the impact of early adversity is conveyed from mother to child. To accomplish these aims this project is led by multiple principal investigators with specific expertise in: (1) mother-infant communication; (2) neurobiological effects of childhood abuse and (3) neonatal neuroimaging.

Copy number variations (CNVs) at the BP4-BP5 segment of chromosome 16p11.2 are known to be associated with neurodevelopmental and psychiatric disorders including language impairment, intellectual disability, ASD, epilepsy, and schizophrenia. The purpose of this study is to detect alterations in sulcal patterns in 16p11.2 deletion and duplication carriers compared to typically developing individuals. We aim to acquire new information from a “genetics first” approach that helps to contribute to the understanding of structural brain abnormalities, which may occur during the early gestational period, and how these abnormalities are related to the 16p11.2 associated neurodevelopmental disorders.

Congenital heart disease (CHD) is one of the most common congenital disorders, affecting about 1% of all live births. More than half of children with moderate or severe CHD have neurodevelopmental disabilities (NDD) that persist into later life. Quantitative methods that can objectively identify subjects at high risk for NDD as early as possible are needed to allow for characterization of the mechanisms underlying NDD and monitor the success of potential therapeutic interventions. Since patients with CHD suffer high rates of NDD and show both severe in utero reduction in oxygenated cerebral blood supply and frequent damaging genetic variants, the goal of this study is to examine sulcal patterns and regional cortical growth (thickness and surface area) to indicate the effects of genetic variants and altered cerebral hemodynamics respectively using a large dataset of retrospective and prospective longitudinal MRIs from the second trimester to birth with three time points. This study also aims to create the model that predicts postnatal preoperative brain abnormalities using deep learning techniques.

Cortical growth and the resulting increased gyration separate humans from other species and are related to our capacity for high-order cognitive functions. Accordingly, it has been a matter of great interest to understand mechanisms of the human cortical growth and folding, such as genetic and environmental effects. As the placenta plays a vital role in shaping the fetal environment and affecting fetal growth through the exchange of oxygen and nutrients, we can hypothesize that placental transport function may be one of the important non-genetic factors that influence spatio-temporal dynamics of early human cortical growth and folding. The proposed project aims to explore the effect of placental transport function on the fetal cortical growth and folding to define non-genetic, potentially modifiable, prenatal environmental factors affecting early human brain development and subsequent postnatal clinical and behavioral outcomes. We propose a monochorionic twin analysis controlling for genetic and maternal factors and usinginnovative magnetic resonance imaging (MRI) and analysis methods to directly quantify fetal cortical development and placental transport function in vivo.

The purpose of this project is to develop fully automatic pipeline for fetal brain MRI processing including automatic brain segmentation; fetal MRI quality assessment; motion correction; fetal brain tissue segmentation; cortical surface extraction; and volume- and surface-based feature extraction. We are using state-of-the-art medical image processing and deep learning techniques for this pipeline development.

Neonatal encephalopathy (NE) due to hypoxia-ischemia occurs in 6/1000 live term births and can have significant neurological consequences. Therapeutic hypothermia (TH), the current standard of care, primarily acts through decreasing cerebral metabolism yet current brain monitoring tools lack the ability to track cerebral metabolism. The goal of this study is to improve bedside monitoring of brain health in newborn babies at risk of brain injuries by using an advanced non-invasive optical imaging technology to optimize treatment for individual neonates and thus help in improving neurodevelopmental outcomes.

The placenta is a highly vascular organ, responsible for the interaction between the mother and the developing fetus. Despite its critical role during pregnancy, the placenta is currently understudied, with no direct diagnostic tool that can monitor placental function. The premise of this project is to address the need for clinically relevant technological advances in MRI for placental imaging, and to that end we aim to develop robust, quantitative measures of regional placental perfusion, intervillous inflow, oxygen response and oxygen state from late 2​nd trimester to term, with safety analysis and feasibility testing that pave the way for extension to late 1​st trimester. After developing the proposed methods, we hope to provide a better way to assess placental health by distinguishing unhealthy placentas from the healthy ones, and therefore help doctors make more informed decisions when managing high risk pregnancies and determining optimal delivery time.

Currently there are no clear strategies to assess when in the progressive evolution of pediatric moyamoya patients would maximally benefit from surgical intervention to prevent ischemic brain injury, resulting in wide variation in practice among even high-volume tertiary referral centers. Advanced MRI techniques have been developed to quantify cerebral hemodynamics that may be useful in assessing brain physiology relevant to clinical management of moyamoya. However, these techniques have yet to be combined and incorporated into clinical practice because their diagnostic accuracy, both in terms of ischemic risk prognosis and prediction of collateral vessel formation after surgery, remains unknown. In this prospective observational study of pediatric patients with moyamoya we will add arterial spin labeling, breath hold cerebrovascular reserve mapping, and 4D flow sequences to clinical MRI studies in order to quantify cerebral hemodynamics. If successful, new strategies to prevent interval stroke could be assessed and likelihood of response to therapy could be used to prioritize patients most likely to respond for earlier surgical interventions. These techniques might also aid in clinical management of other pediatric cerebrovascular conditions such as vein of Galen malformation, congenital heart disease and sickle cell disease.

For children with drug-resistant epilepsy (DRE), brain surgery is the best treatment to stop seizures. The only way to estimate the brain area that is responsible for seizures – and then remove it – is through the use and combination of various diagnostic tools. Among them, electro-encephalography (EEG) is key: surgical planning depends primarily upon the ability to record seizures and visually identify a clear seizure onset on the EEG. Yet, seizures are unpredictable and sometimes difficult to capture and interpret. The goal of this project is to develop a novel multi-feature approach to analyze EEG that exploits and integrates new “invisible” signal patterns, which can inform us on epileptogenicity without relying on frank epileptiform activity. Our proposed methods integrate concepts from cross-frequency coupling, functional connectivity and graph theory.