Mapping Early Brain-Body Interactions: Associations of Fetal Heart Rate Variation with Newborn Brainstem, Hypothalamic, and Dorsal Anterior Cingulate Cortex Functional Connectivity.

The autonomic nervous system (ANS) regulates the body's physiology, including cardiovascular function. As the ANS develops during the second to third trimester, fetal heart rate variability (HRV) increases while fetal heart rate (HR) decreases. In this way, fetal HR and HRV provide an index of fetal ANS development and future neurobehavioral regulation. Fetal HR and HRV have been associated with child language ability and psychomotor development behavior in toddlerhood. However, their associations with postbirth autonomic brain systems, such as the brainstem, hypothalamus, and dorsal anterior cingulate cortex (dACC), have yet to be investigated even though brain pathways involved in autonomic regulation are well established in older individuals. We assessed whether fetal HR and HRV were associated with the brainstem, hypothalamic, and dACC functional connectivity in newborns. Data were obtained from 60 pregnant individuals (ages 14-42) at 24-27 and 34-37 weeks of gestation using a fetal actocardiograph to generate fetal HR and HRV. During natural sleep, their infants (38 males and 22 females) underwent a fMRI scan between 40 and 46 weeks of postmenstrual age. Our findings relate fetal heart indices to brainstem, hypothalamic, and dACC connectivity and reveal connections with widespread brain regions that may support behavioral and emotional regulation. We demonstrated the basic physiologic association between fetal HR indices and lower- and higher-order brain regions involved in regulatory processes. This work provides the foundation for future behavioral or physiological regulation research in fetuses and infants.

Maternal prenatal immune activation associated with brain tissue microstructure and metabolite concentrations in newborn infants.

Importance: Few translational human studies have assessed the association of prenatal maternal immune activation with altered brain development and psychiatric risk in newborn offspring. Objective: To identify the effects of maternal immune activation during the 2nd and 3rd trimesters of pregnancy on newborn brain metabolite concentrations, tissue microstructure, and longitudinal motor development. Design: Prospective longitudinal cohort study conducted from 2012 - 2017. Setting: Columbia University Irving Medical Center and Weill Cornell Medical College. Participants: 76 nulliparous pregnant women, aged 14 to 19 years, were recruited in their 2nd trimester, and their children were followed through 14 months of age. Exposure: Maternal immune activation indexed by maternal interleukin-6 and C-reactive protein in the 2nd and 3rd trimesters of pregnancy. Main Outcomes and Measures: The main outcomes included (1) newborn metabolite concentrations, measured as N-acetylaspartate, creatine, and choline using Magnetic Resonance Spectroscopy; (2) newborn fractional anisotropy and mean diffusivity measured using Diffusion Tensor Imaging; and (3) indices of motor development assessed prenatally and postnatally at ages 4- and 14-months. Results: Maternal interleukin-6 and C-reactive protein levels in the 2nd or 3rd trimester were significantly positively associated with the N-acetylaspartate, creatine, and choline concentrations in the putamen, thalamus, insula, and anterior limb of the internal capsule. Maternal interleukin-6 was associated with fractional anisotropy in the putamen, insula, thalamus, precuneus, and caudate, and with mean diffusivity in the inferior parietal and middle temporal gyrus. C-reactive protein was associated with fractional anisotropy in the thalamus, insula, and putamen. Regional commonalities were found across imaging modalities, though the direction of the associations differed by immune marker. In addition, a significant positive association was observed between offspring motor development and both maternal interleukin-6 and C-reactive protein (in both trimesters) prenatally and 4- and 14-months of age. Conclusions and Relevance: Using a healthy sample, these findings demonstrate that levels of maternal immune activation in mid- to late pregnancy associate with tissue characteristics in newborn brain regions primarily supporting motor integration/coordination and behavioral regulation and may lead to alterations in motor development.

Utilizing maternal prenatal cognition as a predictor of newborn brain measures of intellectual development.

Identifying reliable indicators of cognitive functioning prior to age five has been challenging. Prior studies have shown that maternal cognition, as indexed by intellectual quotient (IQ) and years of education, predict child intelligence at school age. We examined whether maternal full scale IQ, education, and inhibitory control (index of executive function) are associated with newborn brain measures and toddler language outcomes to assess potential indicators of early cognition. We hypothesized that maternal indices of cognition would be associated with brain areas implicated in intelligence in school-age children and adults in the newborn period. Thirty-seven pregnant women and their newborns underwent an MRI scan. T2-weighted images and surface-based morphometric analysis were used to compute local brain volumes in newborn infants. Maternal cognition indices were associated with local brain volumes for infants in the anterior and posterior cingulate, occipital lobe, and pre/postcentral gyrus - regions associated with IQ, executive function, or sensori-motor functions in children and adults. Maternal education and executive function, but not maternal intelligence, were associated with toddler language scores at 12 and 24 months. Newborn brain volumes did not predict language scores. Overall, the pre/postcentral gyrus and occipital lobe may be unique indicators of early intellectual development in the newborn period. Given that maternal executive function as measured by inhibitory control has robust associations with the newborn brain and is objective, brief, and easy to administer, it may be a useful predictor of early developmental and cognitive capacity for young children.

Cross Atlas Remapping via Optimal Transport (CAROT): Creating connectomes for different atlases when raw data is not available.

Open-source, publicly available neuroimaging datasets - whether from large-scale data collection efforts or pooled from multiple smaller studies - offer unprecedented sample sizes and promote generalization efforts. Releasing data can democratize science, increase the replicability of findings, and lead to discoveries. Partly due to patient privacy, computational, and data storage concerns, researchers typically release preprocessed data with the voxelwise time series parcellated into a map of predefined regions, known as an atlas. However, releasing preprocessed data also limits the choices available to the end-user. This is especially true for connectomics, as connectomes created from different atlases are not directly comparable. Since there exist several atlases with no gold standards, it is unrealistic to have processed, open-source data available from all atlases. Together, these limitations directly inhibit the potential benefits of open-source neuroimaging data. To address these limitations, we introduce Cross Atlas Remapping via Optimal Transport (CAROT) to find a mapping between two atlases. This approach allows data processed from one atlas to be directly transformed into a connectome based on another atlas without the need for raw data access. To validate CAROT, we compare reconstructed connectomes against their original counterparts (i.e., connectomes generated directly from an atlas), demonstrate the utility of transformed connectomes in downstream analyses, and show how a connectome-based predictive model can generalize to publicly available data that was processed with different atlases. Overall, CAROT can reconstruct connectomes from an extensive set of atlases - without needing the raw data - allowing already processed connectomes to be easily reused in a wide range of analyses while eliminating redundant processing efforts. We share this tool as both source code and as a stand-alone web application (http://carotproject.com/).

Machine Learning and Prediction in Fetal, Infant, and Toddler Neuroimaging: A Review and Primer.

Predictive models in neuroimaging are increasingly designed with the intent to improve risk stratification and support interventional efforts in psychiatry. Many of these models have been developed in samples of children school-aged or older. Nevertheless, despite growing evidence that altered brain maturation during the fetal, infant, and toddler (FIT) period modulates risk for poor mental health outcomes in childhood, these models are rarely implemented in FIT samples. Applications of predictive modeling in children of these ages provide an opportunity to develop powerful tools for improved characterization of the neural mechanisms underlying development. To facilitate the broader use of predictive models in FIT neuroimaging, we present a brief primer and systematic review on the methods used in current predictive modeling FIT studies. Reflecting on current practices in more than 100 studies conducted over the past decade, we provide an overview of topics, modalities, and methods commonly used in the field and under-researched areas. We then outline ethical and future considerations for neuroimaging researchers interested in predicting health outcomes in early life, including researchers who may be relatively new to either advanced machine learning methods or using FIT data. Altogether, the last decade of FIT research in machine learning has provided a foundation for accelerating the prediction of early-life trajectories across the full spectrum of illness and health.

Predicting the future of neuroimaging predictive models in mental health.

Predictive modeling using neuroimaging data has the potential to improve our understanding of the neurobiology underlying psychiatric disorders and putatively information interventions. Accordingly, there is a plethora of literature reviewing published studies, the mathematics underlying machine learning, and the best practices for using these approaches. As our knowledge of mental health and machine learning continue to evolve, we instead aim to look forward and "predict" topics that we believe will be important in current and future studies. Some of the most discussed topics in machine learning, such as bias and fairness, the handling of dirty data, and interpretable models, may be less familiar to the broader community using neuroimaging-based predictive modeling in psychiatry. In a similar vein, transdiagnostic research and targeting brain-based features for psychiatric intervention are modern topics in psychiatry that predictive models are well-suited to tackle. In this work, we target an audience who is a researcher familiar with the fundamental procedures of machine learning and who wishes to increase their knowledge of ongoing topics in the field. We aim to accelerate the utility and applications of neuroimaging-based predictive models for psychiatric research by highlighting and considering these topics. Furthermore, though not a focus, these ideas generalize to neuroimaging-based predictive modeling in other clinical neurosciences and predictive modeling with different data types (e.g., digital health data).

An ode to fetal, infant, and toddler neuroimaging: Chronicling early clinical to research applications with MRI, and an introduction to an academic society connecting the field.

Fetal, infant, and toddler neuroimaging is commonly thought of as a development of modern times (last two decades). Yet, this field mobilized shortly after the discovery and implementation of MRI technology. Here, we provide a review of the parallel advancements in the fields of fetal, infant, and toddler neuroimaging, noting the shifts from clinical to research use, and the ongoing challenges in this fast-growing field. We chronicle the pioneering science of fetal, infant, and toddler neuroimaging, highlighting the early studies that set the stage for modern advances in imaging during this developmental period, and the large-scale multi-site efforts which ultimately led to the explosion of interest in the field today. Lastly, we consider the growing pains of the community and the need for an academic society that bridges expertise in developmental neuroscience, clinical science, as well as computational and biomedical engineering, to ensure special consideration of the vulnerable mother-offspring dyad (especially during pregnancy), data quality, and image processing tools that are created, rather than adapted, for the young brain.

Out-of-focus brain image detection in serial tissue sections.

BACKGROUND: A large part of image processing workflow in brain imaging is quality control which is typically done visually. One of the most time consuming steps of the quality control process is classifying an image as in-focus or out-of-focus (OOF). NEW METHOD: In this paper we introduce an automated way of identifying OOF brain images from serial tissue sections in large datasets (>1.5 PB). The method utilizes steerable filters (STF) to derive a focus value (FV) for each image. The FV combined with an outlier detection that applies a dynamic threshold allows for the focus classification of the images. RESULTS: The method was tested by comparing the results of our algorithm with a visual inspection of the same images. The results support that the method works extremely well by successfully identifying OOF images within serial tissue sections with a minimal number of false positives. COMPARISON WITH EXISTING METHODS: Our algorithm was also compared to other methods and metrics and successfully tested in different stacks of images consisting solely of simulated OOF images in order to demonstrate the applicability of the method to other large datasets. CONCLUSIONS: We have presented a practical method to distinguish OOF images from large datasets that include serial tissue sections that can be included in an automated pre-processing image analysis pipeline.

An automated method for removal of striping artifacts in fluorescent whole-slide microscopy.

BACKGROUND: Whole slide scanners often acquire images of tissue sections that are larger than their field of view through tile or line-scanning. The subsequently stitched or aligned images can often suffer from imaging artifacts such as horizontal or vertical stripes. These stripes degrade the image quality in fluorescent biological imaging samples and can also limit the accuracy of any subsequent analyses such as cell segmentation. NEW METHOD: We propose a novel data-driven method of removing stripe artifacts in stitched biological images based on the location of the stripes, background modeling, and illumination correction. This method provides an automated way of removing the stripes of an individual image while preserving image details and quality for subsequent analyses. RESULTS: The results were assessed using both qualitative and quantitative metrics and the algorithm has proven very effective in removing the stripe artifacts from hundreds of brain images. COMPARISON WITH EXISTING METHODS: Several metrics were used to quantify the effectiveness of our proposed method compared to other published techniques. Images with simulated artifacts were created so that full-reference metrics could be applied to demonstrate the applicability of the algorithm for a wider variety of illumination profiles. CONCLUSIONS: We describe a data analysis pipeline that allows for automatic removal of stripes caused by line-scanning. Our proposed method can be applied without the need for separate blank field of view images or use of image batches to model the background, so it is suitable for real-time parallel batch processing of large datasets.