Manual blood smear analysis remains the gold standard to diagnose hematological disorders and infections of blood parasites. However, the analysis and interpretation of peripheral blood smears requires expert users, is time consuming, depends on inter-observer variation, and is not compatible with a high-throughput workflow for clinical routine diagnostics (Dunning & Safo, Biotech. Histochem. 2011, 86, 69–75; Pierre, Clin. Lab. Med., 2002, 22, 279–297). Instead, automated hematology analyzers only flag atypical results which provides no clear classification of diseases and require extensive sample preparation. Label-free image analysis of untouched blood cells would reduce pre-analytical efforts and potentially allows characterization of samples with higher information content compared to both smear analysis and conventional automated flow cytometry, as the blood cell morphology is preserved. Furthermore, preclinical research work is in need for non-invasive analysis of e.g. cancer cells or infected cells to support the discovery of new drugs.
We suggest to apply high-throughput and label-free workflows based on digital holographic microscopy for standardizable image analysis relevant for pre- and clinical diagnosis. In the case of parasitic infections, the label-free detection and analysis of malaria parasites has been addressed by various studies (Anand et al., IEEE Photonics J., 2012, 4, 1456-1464; Seo et al., Appl. Phys. Lett., 2014, 104, 1-4; Park et al, PloS One, 2016, 11, 1-19; Ugele et al., Lab Chip, 2018, 18, 1704-1712). The detection of neglected tropical diseases affecting livestock and humans, such as Chagas disease and Leishmaniosis, has not been addressed so far by the community.
Our platform technology is based on a customized differential holographic microscopy setup, which has been previously described (Ugele et al., Lab Chip, 2018, 18, 1704-1712; Ugele et al., Adv. Sci., 2018, 5, 1800761). Reference data sets of clinical leukemic samples, cancer cell cultures in solution, and in vitro cultures of various parasites were collected to understand the translational potential for this methodology. Hydrodynamic and viscoelastic focusing in a microfluidics channel was used for high-throughput imaging and enrichment/depletion of cell populations without the need for any autofocusing procedures. Morphological parameters describing the inner consistency were calculated from segmented phase images of the cells/parasites and combined with machine learning algorithms for improved analysis by the discovery of label-free biomarkers. In this way, improved subtyping of acute and chronic leukemias, myeloproliferative neoplasms, and further hematological disorders was achieved. Second, a detection of Trypanosoma and Leishmania parasites could be shown and in vitro cultures of Schistosomia mansoni were classified according to different viability stages. Third, the capability of anti-cancer drug candidate screening was demonstrated by monitoring the mesenchymal-epithelial transition of pancreas cancer cell cultures. We envision, that our platform technology has the potential as a cost-efficient method for automated diagnosis of various hematological disorders, parasitic infections, drug screening and monitoring of therapy efficacy. With further integration effort we also believe that the technology can be applied in resource limited settings.