No abstract

Pulmonary tuberculosis (TB) is caused by Mycobacterium tuberculosis in susceptible humans. Here, we infected Diversity Outbred (DO) mice with ∼100 bacilli by aerosol to model responses in a highly heterogeneous population. Following infection, ‘supersusceptible’, ‘susceptible’ and ‘resistant’ phenotypes emerged. TB disease (reduced survival, weight loss, high bacterial load) correlated strongly with neutrophils, neutrophil chemokines, tumor necrosis factor (TNF) and cell death. By contrast, immune cytokines were weak correlates of disease. We next applied statistical and machine learning approaches to our dataset of cytokines and chemokines from lungs and blood. Six molecules from the lung: TNF, CXCL1, CXCL2, CXCL5, interferon-γ (IFN-γ), interleukin 12 (IL-12); and two molecules from blood – IL-2 and TNF – were identified as being important by applying both statistical and machine learning methods. Using molecular features to generate tree classifiers, CXCL1, CXCL2 and CXCL5 distinguished four classes (supersusceptible, susceptible, resistant and non-infected) from each other with approximately 77% accuracy using completely independent experimental data. By contrast, models based on other molecules were less accurate. Low to no IFN-γ, IL-12, IL-2 and IL-10 successfully discriminated non-infected mice from infected mice but failed to discriminate disease status amongst supersusceptible, susceptible and resistant M.-tuberculosis-infected DO mice. Additional analyses identified CXCL1 as a promising peripheral biomarker of disease and of CXCL1 production in the lungs. From these results, we conclude that: (1) DO mice respond variably to M. tuberculosis infection and will be useful to identify pathways involving necrosis and neutrophils; (2) data from DO mice is suited for machine learning methods to build, validate and test models with independent data based solely on molecular biomarkers; (3) low levels of immunological cytokines best indicate a lack of exposure to M. tuberculosis but cannot distinguish infection from disease.

The development of whole slide scanners has revolutionized the field of digital pathology. Unfortunately, whole slide scanners often produce images with out-of-focus/blurry areas that limit the amount of tissue available for a pathologist to make accurate diagnosis/prognosis. Moreover, these artifacts hamper the performance of computerized image analysis systems. These areas are typically identified by visual inspection, which leads to a subjective evaluation causing high intra- and inter-observer variability. Moreover, this process is both tedious, and time-consuming. The aim of this study is to develop a deep learning based software called, DeepFocus, which can automatically detect and segment blurry areas in digital whole slide images to address these problems. DeepFocus is built on TensorFlow, an open source library that exploits data flow graphs for efficient numerical computation. DeepFocus was trained by using 16 different H&E and IHC-stained slides that were systematically scanned on nine different focal planes, generating 216,000 samples with varying amounts of blurriness. When trained and tested on two independent datasets, DeepFocus resulted in an average accuracy of 93.2% (± 9.6%), which is a 23.8% improvement over an existing method. DeepFocus has the potential to be integrated with whole slide scanners to automatically re-scan problematic areas, hence improving the overall image quality for pathologists and image analysis algorithms.

Background: Identifying which individuals will develop tuberculosis (TB) remains an unresolved problem due to few animal models and computational approaches that effectively address its heterogeneity. To meet these shortcomings, we show that Diversity Outbred (DO) mice reflect human-like genetic diversity and develop human-like lung granulomas when infected with Mycobacterium tuberculosis (M.tb). Methods: Following M.tb infection, a "supersusceptible" phenotype develops in approximately one-third of DO mice characterized by rapid morbidity and mortality within 8 weeks. These supersusceptible DO mice develop lung granulomas patterns akin to humans. This led us to utilize deep learning to identify supersusceptibility from hematoxylin & eosin (H&E) lung tissue sections utilizing only clinical outcomes (supersusceptible or not-supersusceptible) as labels. Findings: The proposed machine learning model diagnosed supersusceptibility with high accuracy (91.50 § 4.68%) compared to two expert pathologists using H&E stained lung sections (94.95% and 94.58%). Two nonexperts used the imaging biomarker to diagnose supersusceptibility with high accuracy (88.25% and 87.95%) and agreement (96.00%). A board-certified veterinary pathologist (GB) examined the imaging biomarker and determined the model was making diagnostic decisions using a form of granuloma necrosis (karyorrhectic and pyknotic nuclear debris). This was corroborated by one other board-certified veterinary pathologist. Finally, the imaging biomarker was quantified, providing a novel means to convert visual patterns within granulomas to data suitable for statistical analyses. Implications: Overall, our results have translatable implication to improve our understanding of TB and also to the broader field of computational pathology in which clinical outcomes alone can drive automatic identification of interpretable imaging biomarkers, knowledge discovery, and validation of existing clinical biomarkers.

No abstract