Artificial intelligence (AI) continues to garner substantial interest in medical imaging. The potential applications are vast and include the entirety of the medical imaging life cycle from image creation to diagnosis to outcome prediction. The chief obstacles to development and clinical implementation of AI algorithms include availability of sufficiently large, curated, and representative training data that includes expert labeling (eg, annotations). Current supervised AI methods require a curation process for data to optimally train, validate, and test algorithms. Currently, most research groups and industry have limited data access based on small sample sizes from small geographic areas. In addition, the preparation of data is a costly and time-intensive process, the results of which are algorithms with limited utility and poor generalization. In this article, the authors describe fundamental steps for preparing medical imaging data in AI algorithm development, explain current limitations to data curation, and explore new approaches to address the problem of data availability.

The Library of Integrated Network-Based Cellular Signatures (LINCS) is an NIH Common Fund program that catalogs how human cells globally respond to chemical, genetic, and disease perturbations. Resources generated by LINCS include experimental and computational methods, visualization tools, molecular and imaging data, and signatures. By assembling an integrated picture of the range of responses of human cells exposed to many perturbations, the LINCS program aims to better understand human disease and to advance the development of new therapies. Perturbations under study include drugs, genetic perturbations, tissue micro-environments, antibodies, and disease-causing mutations. Responses to perturbations are measured by transcript profiling, mass spectrometry, cell imaging, and biochemical methods, among other assays. The LINCS program focuses on cellular physiology shared among tissues and cell types relevant to an array of diseases, including cancer, heart disease, and neurodegenerative disorders. This Perspective describes LINCS technologies, datasets, tools, and approaches to data accessibility and reusability.

SUMMARYBrain microvascular endothelial cells (BMECs) are an essential component of the blood-brain barrier (BBB) that shields the brain against toxins and immune cells. While BBB dysfunction exists in neurological disorders, including Huntington’s disease (HD), it is not known if BMECs themselves are functionally compromised to promote BBB dysfunction. Further, the underlying mechanisms of BBB dysfunction remain elusive given limitations with mouse models and post-mortem tissue to identify primary deficits. We undertook a transcriptome and functional analysis of human induced pluripotent stem cell (iPSC)-derived BMECs (iBMEC) from HD patients or unaffected controls. We demonstrate that HD iBMECs have intrinsic abnormalities in angiogenesis and barrier properties, as well as in signaling pathways governing these processes. Thus, our findings provide an iPSC-derived BBB model for a neurodegenerative disease and demonstrate autonomous neurovascular deficits that may underlie HD pathology with implications for therapeutics and drug delivery.

Inactivating mutations in the thyroid hormone (TH) transporter Monocarboxylate transporter 8 (MCT8) cause severe psychomotor retardation in children. Animal models do not reflect the biology of the human disease. Using patient-specific induced pluripotent stem cells (iPSCs), we generated MCT8-deficient neural cells that showed normal TH-dependent neuronal properties and maturation. However, the blood-brain barrier (BBB) controls TH entry into the brain, and reduced TH availability to neural cells could instead underlie the diseased phenotype. To test potential BBB involvement, we generated an iPSC-based BBB model of MCT8 deficiency, and we found that MCT8 was necessary for polarized influx of the active form of TH across the BBB. We also found that a candidate drug did not appreciably cross the mutant BBB. Our results therefore clarify the underlying physiological basis of this disorder, and they suggest that circumventing the diseased BBB to deliver active TH to the brain could be a viable therapeutic strategy.

SummarySocial gliding motility in Myxococcus xanthus depends on the presence of Type IV pili. To begin to examine the role of pili in social motility, 17 mutants were identified which had lost social motility, but still expressed pili. Four of these mutants carry point mutations which mapped to a locus upstream of the recently identified pilS, pilR, and pilA genes. Sequencing of this locus revealed a gene with homology to pilT from Pseudomonas aeruginosa. Sequencing of the four point mutations revealed that they occurred within the M. xanthus pilT locus. A markerless deletion within M. xanthus pilT, similar to the four point mutations, disrupted social gliding behaviour but did not interfere with pilus formation or pilus-dependent cellcell agglutination. Using time-lapse videomicroscopy, residual social motility was observed in dsp ¹ strains (known to be deficient in fibril but not pilus production); this was not observed in a ⌬pilT dsp ¹ double mutant. Two genes flanking pilT were also sequenced, and found to have homology to pilB and pilC from P. aeruginosa. Markerless deletions within these genes caused both pilus and social-motility defects. These results indicate that M. xanthus pilB and pilC are required for pilus biogenesis, while pilT is required for assembled pili to play their role in social motility. Thus, pilB, pilT, pilC, pilS, pilR and pilA form a contiguous cluster of pil genes required for social motility.