|
Almost 70 percent of all data generated in life sciences is in the form of images, but the potential use of this data is only partially tapped. The latest high-throughput image-acquisition devices in the laboratory produce thousands of images a day and volumes are increasing. However, the utility of image analysis has long been mitigated by high volumes of unprocessed data. The process of manually analyzing images is labor intensive and inherently subjective.
While no system can supplant trained specialists, adopting systems to help bring objectivity and automation to image analysis can help streamline research. The challenge is many current image-analysis software tools require extensive programming experience, time-intensive training and have inaccessible interfaces.
Despite years of research and development, automated image-analysis software has fallen short of its promise, and proliferating image-acquisition systems add a layer of complexity. Currently, bioscience and pharmaceutical companies use more than 100 different image-acquisition devices. Accompanying image-analysis software is often purchased as point solutions to solve specific problems, or is embedded in the image-acquisition devices themselves. Although image-analysis software may improve the speed and accuracy of analysis, many accept only particular inputs or file types, restricting their usefulness. As translational research becomes increasingly important, this challenge presents major hurdles to implementation.
Translational Research
As defined by the National Cancer Institute, "Translational research transforms scientific discoveries arising from laboratory, clinical or population studies into clinical applications to reduce cancer incidence, morbidity and mortality," (National Cancer Institute. TRWG definition of translational research. Available at: http://www.cancer.gov/trwg/TRWG-definition-and-TR-continuum. Last accessed Oct. 19, 2009) Biomarkers discovered and validated in translational research are utilized for disease prognostication, clinical trial enrichment,
 |
| ADVANCE thanks Definiens |
companion assays for targeted therapies and as surrogates for toxicity.
Biopharmaceutical companies are prioritizing their drug development efforts by using biomarkers identified and validated through translational research, and this trend is likely to increase. To be successful in both translational research and drug development, scientists need to compare pathological reactions in various animal and human models, all of which are imaged by multiple devices.
Additionally, molecular imaging and cell-based assays give early indications in pre-clinical research of the effectiveness and toxicity of therapeutics. Images acquired from tissue samples and non-invasive scans (CT, MRI) reveal valuable insights about a disease, its progression and whether or not a patient is responding to treatment. With various imaging modalities involved, there is a need for accessible, automated image-analysis solutions capable of processing and comparing image data from a multitude of acquisitions and image formats.
Cognitive Approach
Definiens Cognition Network Technology is an advanced approach to extracting intelligence from images that emulates human cognitive processes. In contrast to traditional pixel-based or pattern-recognition methodologies, Definiens' segmentation and classification processes examine image objects in relationship to each other. In this way, Definiens software not only recognizes patterns, but more importantly, accurately segments images in which there are no obvious or reproducible patterns; a common occurrence when assimilating information from complex, heterogeneous biological systems.
The data resulting from Definiens' image analysis is multi-parametric, providing scientists with a comprehensive, quantitative understanding of complex images. The technology is employed in cell-based assays, histology specimens and medical imaging scans; in 2-D, 3-D and 4-D.
The recently introduced Definiens Tissue Studio software application is built upon Definiens Cognition Network Technology. Designed for pathologists, it supports a wide array of immunohistochemistry, hematoxylin & eosin and immunofluorescence stains, and is able to identify and analyze heterogeneous tissues and the cells and sub-cellular objects within them.
 The application was developed iteratively to address pathologists' needs in translational research of tissue-based biomarkers. During the development of Definiens Tissue Studio, particular attention was given to designing an accessible, simplified interface. Pathologists without extensive computer programming expertise are able to use the "paint brush" tool to teach the application to demarcate regions of interest (ROI). The application can be "trained" to detect regions of interest and enables users to work with up to four images simultaneously, allowing for rapid identification and quantification of biomarkers in cellular and sub-cellular compartments. The application can then be used in "batch mode" for automated parallel processing.
As imaging data becomes an integral part of translational research, there will be a need for image-analysis systems that are intuitive and accessible to researchers, while also capable of handling input from multiple modalities. With this in mind, Definiens has developed image-analysis applications for medical imaging and life sciences to enable researchers to process imaging data from all common imaging modalities. Providing image-analysis software solutions enabling researchers and clinicians to work more efficiently and which provide more complete insight into complex biological systems supports translational research and the drive toward personalized medicine.
Peter A. Duncan is director, Marketing and Business Development, Life Sciences, Parsippany, NJ.
|