Development and Dissemination of MuscleMiner: An Imaging Informatics Tool for Muscle

  • Esser, Karyn (PI)

Grants and Contracts Details

Description

Profound skeletal muscle weakness, atrophy, fatigue, and dysfunction are features of a variety of chronic diseases, as well as overt muscle pathologies. Muscle biological research and clinical diagnosis routinely involve histochemical analyses of muscle biopsies, in addition to more specific histochemical and immunohistochemical antigen detection and localization. There is agreement that important morphometric characteristics of muscle fibers, such as fiber area, the number and position of myonuclei, cellular infiltration and fibrosis, etc., are critical factors that determine the health and function (e.g. quality) of the muscle. Although these features are extremely important, currently, quantification of muscle characteristics in microscopic images is still a manual or, at best, a semiautomatic process. This labor-intensive process is not only time consuming but also prone to errors, leading to high inter-observer variability. Basic scientists often need to spend months of valuable time to measure one single morphometric feature, such as cross-sectional area. In clinical practice, subtle changes in morphometric parameters, which are not always apparent or quantifiable through visual inspections but can provide critical diagnostic and prognostic clues, have never been rigorously measured for muscle diseases. When muscle characteristics are calculated by computer-aided analysis, objectivity, accuracy, and reproducibility improve significantly. The lack of computational tool is a barrier not only to the rapid progress of image acquisition technology such as whole slide scanning in basic science research, but also to the increasing clinical needs for objective and reproducible diagnoses and prognoses. Beyond traditional biomedical informatics that has already attracted a lot of research attentions and tool development effort, a recent upsurge in worldwide attention on imaging informatics for analyzing and retrieving microscopic image data is emerging rapidly. MuscleMiner is proposed to fill the gap by developing an objective analytical tool and an efficient imaging informatics system for muscle image archiving, managing, retrieving, visualizing, and analyzing. The objectives of this proposal are: 1) Design, develop, and validate a robust image analysis and imaging informatics system, MuscleMiner, for basic science and clinical muscle research; 2) Develop advanced computational and learning algorithms to discover novel image markers; 3) Build a Cloud-computing unit to support high throughput, large scale processing of the muscle images.
StatusFinished
Effective start/end date9/8/147/31/15

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.