In her presentation, Marta Czapranska, Senior Scientist at Spatial Biology, Concept Life Sciences, discussed innovative approaches to overcoming scientific and technical challenges in digital pathology. Czapranska began by introducing Concept Life Sciences as a contract research organisation specialising in the journey from concept to clinic, with capabilities spanning biology, chemistry, drug morphology, and pharmacokinetics. The company operates across several UK sites, with Czapranska’s team based in Edinburgh, focusing on spatial biology and delivering comprehensive, up-to-date processes, from tissue sourcing to image analysis.
Czapranska highlighted the group’s commitment to bespoke solutions, particularly in image analysis using Visiopharm software. She described a recent project where the team was tasked with determining whether a drug affected collagen fibre orientation in the upper dermis. Rather than designing a costly new study, the team leveraged existing samples and analytical tools. They developed a method using chromacity red and polylinear local linear transformation to detect and measure fibre orientation, referencing the epidermis as a baseline. Although the study did not reveal significant differences between drug responders and non-responders, the approach was well received by the client for its efficiency and cost-effectiveness. Czapranska noted that further improvements, such as segmenting lesions into smaller regions and increasing sample size, could enhance the method’s sensitivity.
The presentation then shifted to the importance of high-quality staining in image analysis, particularly for biomarker discovery. Czapranska described challenges encountered with variable staining intensity across samples and explained how the team developed a deep learning application to detect marker features based on both intensity and staining patterns. This approach, combined with Phenoplex software, allowed for more accurate and scalable analysis, reducing the time and manual adjustment required compared to traditional methods. The deep learning method also enabled the creation of a reusable library of detection applications, streamlining future studies.
Czapranska concluded by emphasising Concept Life Sciences’ dedication to robust, reproducible, and high-precision spatial biology data, supporting biomarker discovery and translational research through a blend of academic insight and industry expertise.
