Definiens, a healthcare company that advances personalized medicine through image analysis and digital pathology solutions, has released results from an industry survey on the drivers of automated image analysis adoption and predictions on the future of digital pathology. Findings showed that scientists and pathologists are adopting automated image analysis solutions primarily to obtain greater confidence in their data and to address the increasing volume of images that require detailed analyses. The use of image analysis in this manner is an enabler of tissue phenomics, or the ability to quantify all relevant morphological features on a tissue slide.
The 2013 automated image analysis adoption survey conducted by Definiens was an anonymous electronic survey sent to scientists and pathologists in the pharmaceutical, biotech and academic industries.
"This survey confirmed that with the large number of digital tissue images being produced using whole slide image scanners, the need for automated image analysis is becoming much more prevalent. By using image analysis for the datafication of tissue, scientists and pathologists can achieve the confidence levels necessary to make critical decisions in drug discovery or biomarker development," said Thomas Heydler, CEO of Definiens.
The study found the top four drivers of image analysis adoption, in order of importance, are:
1. Gain greater confidence in data and conclusions. The decisions being made utilizing digital images are highly important. They range from discovering potential targets for further study to go/no-go decisions on potential drug targets and biomarkers for companion diagnostics. Automated image analysis strengthens the confidence in results by establishing a set standard for measurement that scientists and pathologists can apply to all images in a study cohort.
2. Address the volume of images requiring analysis. Digital image scanners enable the creation of hundreds to thousands of images in a record amount of time. Those images require analysis, and manually reviewing them can take a significant amount of time scientists and pathologists don't have. Developing automated algorithms to independently analyze all images eliminates tedious work and while ensuring all information in the tissue can be utilized by the pathologist.
3. Correlate image data with other data sources. Unlike manual analysis, where tissue image data is reduced to qualitative scores, automated image analysis enables the information in tissues to be turned into quantifiable, discrete data points. The tissue data output can be easily aligned and compared with data points from genomic analysis or patient outcomes to discover and develop disease biomarkers in ways single sources of data alone don't allow.
4. Quantify complex tissue features and morphology. Automated image analysis is not limited to single data points. Scientists and pathologists can now ask complex questions and develop algorithms that look at and quantify features and morphology the human eye can't manage on a large scale.
To see a full summary of the survey results, including results on why users of digital pathology have not adopted image analysis, please visit the Definiens website.