Image Mining for Comparative Analysis of Expression Patterns in Tissue Microarrays

The capacity to distinguish among subclasses of pathology affects how aggressively patients are treated, which medications are appropriate, and what levels of risk are justified. Current therapies and treatment regimens are often based upon classification strategies which are limited in terms of their capacity to discriminate among specific tumor groups. Recent technological advances make it possible to analyze tumors at the genomic, RNA, and protein expression level. Tissue microarrays (TMA) enable investigators to confirm clinico-pathologic correlations which have been identified in whole histology slide tissue sections. Unfortunately, one of the primary contributing factors leading to inconsistencies during the evaluation process arises from the inherent, subjective impressions of the individuals performing the assessment. The literature, including several systematic studies conducted by our team, shows that when characterizations are based upon computer-aided analysis, objectivity, reproducibility and sensitivity improve considerably. Advanced imaging and computational tools could potentially enable investigators to detect and track subtle changes in measurable parameters leading to the discovery of novel diagnostic and prognostic clues which are not apparent by human visual inspection alone.