MOTIVATION: The molecular subtyping of gastric cancer (adenocarcinoma) into four main subtypes based on integrated multiomics profiles, as proposed by The Cancer Genome Atlas (TCGA) initiative, represents an effective strategy for patient stratification. However, this approach requires the use of multiple technological platforms, and is quite expensive and time-consuming to perform. A computational approach that uses histopathological image data to infer molecular subtypes could be a practical, cost- and time-efficient complementary tool for prognostic and clinical management purposes. RESULTS: Here, we propose a deep learning ensemble approach (called DEMoS) capable of predicting the four recognized molecular subtypes of gastric cancer directly from histopathological images. DEMoS achieved tile-level area under the receiver-operating characteristic curve (AUROC) values of 0.785, 0.668, 0.762 and 0.811 for the prediction of these four subtypes of gastric cancer [i.e. (i) Epstein-Barr (EBV)-infected, (ii) microsatellite instability (MSI), (iii) genomically stable (GS) and (iv) chromosomally unstable tumors (CIN)] using an independent test dataset, respectively. At the patient-level, it achieved AUROC values of 0.897, 0.764, 0.890 and 0.898, respectively. Thus, these four subtypes are well-predicted by DEMoS. Benchmarking experiments further suggest that DEMoS is able to achieve an improved classification performance for image-based subtyping and prevent model overfitting. This study highlights the feasibility of using a deep learning ensemble-based method to rapidly and reliably subtype gastric cancer (adenocarcinoma) solely using features from histopathological images. AVAILABILITY AND IMPLEMENTATION: All whole slide images used in this study was collected from the TCGA database. This study builds upon our previously published HEAL framework, with related documentation and tutorials available at http://heal.erc.monash.edu.au. The source code and related models are freely accessible at https://github.com/Docurdt/DEMoS.git. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Digital pathology supports analysis of histopathological images using deep learning methods at a large-scale. However, applications of deep learning in this area have been limited by the complexities of configuration of the computational environment and of hyperparameter optimization, which hinder deployment and reduce reproducibility. RESULTS: Here, we propose HEAL, a deep learning-based automated framework for easy, flexible and multi-faceted histopathological image analysis. We demonstrate its utility and functionality by performing two case studies on lung cancer and one on colon cancer. Leveraging the capability of Docker, HEAL represents an ideal end-to-end tool to conduct complex histopathological analysis and enables deep learning in a broad range of applications for cancer image analysis. AVAILABILITY AND IMPLEMENTATION: The docker image of HEAL is available at https://hub.docker.com/r/docurdt/heal and related documentation and datasets are available at http://heal.erc.monash.edu.au. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Outline font technology has long been established as the standard way to represent typefaces, allowing characters to be represented independently of print size and resolution. Although outline font technologies are mature and produce results of sufficient quality for professional printing applications, they are inherently inflexible, which presents limitations in a number of document engineering applications. In the 1990s, the topic of finding a successor to outline fonts was a hot topic of research. Unfortunately, none of the methods developed at the time were successful in replacing outline font technology and this field of research has since then declined sharply in popularity.

A few years ago, a perceptually-tuned grayscale character generation technique was developed in order to automatically synthesize grayscale characters looking like manually-tuned pixmap characters. Weight and contrast controlled grayscale characters are obtained by grid-fitting the scaled character contours. However, this technique requires that hinting information be added to the outline font description. Adding hinting information to each outline character generally requires a considerable amount of human intervention. Our component-based parametrisable font system is a newly developed font description and reproduction technology. It incorporates for each basic character shape a software method responsible for the synthesis of an instance of that character. A given font is synthesized by providing appropriate font parameters to these character synthesis methods. Numerous concrete fonts can be derived by simply varying the parameters. Such variations offer high flexibility for synthesizing derived fonts (variations in condensation, weight and contrast) and enable saving a considerable amount of storage space. This paper shows that with component-based parametrisable fonts, high quality perceptually-tuned grayscale characters can be generated without requiring hinting information. Generating perceptually-tuned grayscale characters with parametrized component-based fonts consists in automatically adapting the phase of some of the character's parameters in respect to the underlying grid and in ensuring that thin character parts are strong enough not to disappear (weight-control).