Prostate cancer diagnosis has historically relied on prostate-specific antigen (PSA) screening, but accumulating evidence indicates that PSA testing is associated with over-treatment. As a result, it cannot be reliably used to discriminate potentially lethal tumours from indolent (non-aggressive) disease. Histopathology can also be used to diagnose prostate cancer, but it is a low-throughput, labour-intensive technique that involves manual scoring of tissue slides. A pathologist usually performs microscopic examination of prostate tissue and cellular components to grade the level of disease progression. Recently, in both the academic and industrial sectors, image libraries containing digital images of prostate tissue slides and pathologist scores have been created. Streamlining this process requires tools for data integration and mining. Towards this aim, the EU-funded FAST-PATH (Fast-tracking pathology via automated image analysis and high-performance computing: Application to prostate cancer diagnostics) project set out to develop novel tissue imaging technologies. The industrial partners of FAST-PATH validated novel prostate biomarker imaging algorithms and image search techniques that could be utilised for prostate cancer diagnosis. Substantial improvements and refinements to proprietary digital image analysis software led to the quantification of specific prostate cancer biomarkers expressed in the nucleus, membrane and cytoplasm of cells. This quantitative immunohistochemistry tool, used alongside computer-based tissue imaging, facilitated the high-throughput diagnostic analysis of prostate core biopsies. In addition, one of the small and medium enterprises in FAST-PATH commercialised a web-based product called TissueMark, which enabled automation and incorporated digital pathology into routine assessment of tissue samples. TissueMark won the 2014 Frost & Sullivan European award for new product innovation. Collectively, the FAST-PATH automated digital tools comprise a significant improvement to existing methods. Their implementation into routine histopathology screening is expected to expedite and standardise diagnostic processes.
Cancer, diagnosis, prostate cancer, histopathology, imaging, biomarker