Accurate mapping of normal and abnormal patterns of brain development in fetuses and premature neonates is a key factor in the early detection of developmental disorders as well as understanding how external factors can influence early brain growth. The combination of fast multi-slice MRI techniques with computer vision algorithms has recently provided the ability to reliably image 3D brain structure in-utero providing valuable information inaccessible to ultrasound imaging. To complement this new structural information, recently there has been increasing interest in the use of functional MR imaging (fMRI) to map the development of resting state brain function in children, premature neonates and, most recently, fetuses in-utero. This work has revealed signs of a predictable developmental path in the formation of resting state patterns of brain activity present in adults and abnormalities in these patterns in children have been linked with neurological and neuropsychiatric conditions in later life. Such fMRI methods, previously used in adults and children, could provide a unique new window into functional activity in the developing fetal brain. However, the imaging techniques required suffer from an important limitation for use in the fetus: they make use of repeated acquisitions where subtle changes in MR signal provide the measure of interest. Fetal head motion within the scanner perturbs both measurement location and signal level due to the changing relationship between fetal anatomy, maternal anatomy and the scanner. Based on our preliminary results, in this proposal we plan to develop a set of new signal and geometry correction methods specifically for fetal fMRI that combine novel acquisition techniques with post-processing algorithms to provide a new route to addressing these unique problems. This will allow us to collect the first accurate functional MRI data from a range of un-sedated fetuses in ages and activity levels seen in clinical studies. We will develop and validate these methods together with complementary pattern analysis techniques specifically aimed at motion scattered data and employ them to build the first combined 4D structure-function map of brain development in-utero. This will show for the first time the temporal and spatial relationship between structural changes and the development of resting state patterns of brain activity covering the critical age of first clinical MRI scan and the following period of cortical folding. his will help answer such questions as: which tissue zones of the fetal brain such activity begins, and whether the functional patterns are related to the timing of the formation of specific cortical folds. We will collect a range of data that captures fetuses both at different ages and different states of activity representing those seen in typical clinical studies, and in addition collect valuable outcome measures on the babies after birth against which in-utero measure will be evaluated. We release the data to the community as a whole in the form of an open-access 4D atlas. This combined structure-function data will provide a unique new reference for both neuroscience and, in the longer term, clinical evaluation of brain health during pregnancy and premature birth.
Clinically, improved in-utero evaluation of the fetal brain is a key concern for obstetricians and pediatricians in managing complex pregnancies and there is also now an increasing public health awareness of the influence of the in-utero environment, in terms of factors such as stress and diet, on long-term health in adult life. Functional connectivit imaging of the brain in childhood has revealed the presence of early markers of later cognitive and neuropsychological problems. The ability to map these same properties in utero, promises to provide a rich set of very specific early markers that can be used to understand the impact of the in-utero environment on brain function and feasibly provide a route to the use of early neuro-protective agents and procedures early in childhood.
Showing the most recent 10 out of 11 publications
