How Machine Learning is Powering Neuroimaging to Improve Brain Health.

This report presents an overview of how machine learning is rapidly advancing clinical translational imaging in ways that will aid in the early detection, prediction, and treatment of diseases that threaten brain health. Towards this goal, we aresharing the information presented at a symposium, "Neuroimaging Indicators of Brain Structure and Function - Closing the Gap Between Research and Clinical Application", co-hosted by the McCance Center for Brain Health at Mass General Hospital and the MIT HST Neuroimaging Training Program on February 12, 2021. The symposium focused on the potential for machine learning approaches, applied to increasingly large-scale neuroimaging datasets, to transform healthcare delivery and change the trajectory of brain health by addressing brain care earlier in the lifespan. While not exhaustive, this overview uniquely addresses many of the technical challenges from image formation, to analysis and visualization, to synthesis and incorporation into the clinical workflow. Some of the ethical challenges inherent to this work are also explored, as are some of the regulatory requirements for implementation. We seek to educate, motivate, and inspire graduate students, postdoctoral fellows, and early career investigators to contribute to a future where neuroimaging meaningfully contributes to the maintenance of brain health.

Mineral Distribution Spatially Patterns Bone Marrow Stromal Cell Behavior on Monolithic Bone Scaffolds.

Interfaces between soft tissue and bone are characterized by transitional gradients in composition and structure that mediate substantial changes in mechanical properties. For interfacial tissue engineering, scaffolds with mineral gradients have shown promise in controlling osteogenic behavior of seeded bone marrow stromal cells (bMSCs). Previously, we have demonstrated a 'top-down' method for creating monolithic bone-derived scaffolds with patterned mineral distributions similar to native tissue. In the present work, we evaluated the ability of these scaffolds to pattern osteogenic behavior in bMSCs in basic, osteogenic, and chondrogenic biochemical environments. Immunohistochemical (IHC) and histological stains were used to characterize cellular behavior as a function of local mineral content. Alkaline phosphatase, an early marker of osteogenesis, and osteocalcin, a late marker of osteogenesis, were positively correlated with mineral content in basic, osteogenic, and chondrogenic media. The difference in bMSC behavior between the mineralized and demineralized regions was most pronounced in an basic biochemical environment. In the mineralized regions of the scaffold, osteogenic markers were clearly present as early as 4 days in culture. In osteogenic media, osteogenic behavior was observed across the entire scaffold, whereas in chondrogenic media, there was an overall reduction in osteogenic biomarkers. Overall, these results indicate local mineral content of the scaffold plays a key role in spatially patterning bMSC behavior. Our results can be utilized for the development of interfacial tissue engineered scaffolds and understanding the role of local environment in determining bMSC behavior. STATEMENT OF SIGNIFICANCE: Soft tissue-to-bone interfaces, such as tendon-bone, ligament-bone, and cartilage-bone, are ubiquitous in mammalian musculoskeletal systems. These interfacial tissues have distinct, hierarchically-structured gradients of cellular, biochemical, and materials components. Given the complexity of the biological structures, interfacial tissues present unique challenges for tissue engineering. Here, we demonstrate that material-derived cues can spatially pattern osteogenic behavior in bone marrow stromal cells (bMSCs). Specifically, we observed that when the bMSCs are cultured on bone-derived scaffolds with mineral gradients, cells in contact with higher mineral content display osteogenic behavior at earlier times than those on the unmineralized substrate. The ability to pattern the cellular complexity found in native interfaces while maintaining biologically relevant structures is a key step towards creating engineered tissue interfaces.

Top-down Fabrication of Spatially Controlled Mineral-Gradient Scaffolds for Interfacial Tissue Engineering.

Materials engineering can generally be divided into "bottom-up" and "top-down" approaches, where current state-of-the-art methodologies are bottom-up, relying on the advent of atomic-scale technologies. Applying bottom-up approaches to biological tissues is challenging due to the inherent complexity of these systems. Top-down methodologies provide many advantages over bottom-up approaches for biological tissues, given that some of the complexity is already built into the system. Here, we generate interfacial scaffolds by the spatially controlled removal of mineral content from trabecular bone using a chelating solution. We controlled the degree and location of the mineral interface, producing scaffolds that support cell growth, while maintaining the hierarchical structure of these tissues. We characterized the structural and compositional gradients across the scaffold using X-ray diffraction, microcomputed tomography (µCT), and Raman microscopy, revealing the presence of mineral gradients on the scale of 20 - 40 µm. Using these data, we generated a model showing the dependence of mineral removal as function of time in the chelating solution and initial bone morphology, specifically trabecular density. These scaffolds will be useful for interfacial tissue engineering, with application in the fields of orthopedics, developmental biology, and cancer metastasis to bone.