Label-Free Analysis and Live Sorting of Functional Cell States by Multi-Angle Pulse Shape Flow Cytometry

Reference Number TO 22-00033

Invention Novelty

MAPS-FC captures the complete, time-resolved shape of scattered-light pulses from individual cells across multiple angular directions and analyses these waveforms directly with neural networks. Unlike conventional cytometers, which reduce scatter to low-dimensional, angle-integrated FSC/SSC values, and unlike imaging-based platforms, which reconstruct cell images, MAPS-FC extracts biological information from the raw pulse dynamics themselves – without labels, image reconstruction or segmentation.

Value Proposition

Functional cell states such as cell-cycle phase are conventionally accessible only through labelling that fixes, permeabilizes or otherwise alters the cell, precluding further use. MAPS-FC identifies such states from intrinsic optical properties alone and leaves cells viable and sortable for single-cell sequencing, culture or functional retesting. This reduces panel development effort and opens access to populations for which no robust surface marker exists.

Label-free classification and sorting of functional cell states based on multi-angle scattered-light pulse profiles and neural-network analysis.

Technology Description

Cells passing a focused laser beam generate scattered-light pulses that are recorded by a fibre array at multiple forward-scatter angles. For each event and each angle, the full time-dependent waveform is digitised. These multidimensional signatures are analysed by wavelet-based transforms and, in the current generation, by trained neural networks including recurrent architectures that resolve subtle temporal features of the pulse. Classification results can be coupled to sorting decisions. The approach is compatible with fluorescence detection and with established flow cytometry workflows.

Commercial Opportunity

MAPS-FC closes a gap between fluorescence cytometry, imaging-based systems and downstream omics. First target market is research use in preclinical immunology, inflammation, immuno-oncology, with pharma and biotech R&D groups and translational academic centers as first customers. We are seeking partners for co-development / validation of specific use cases, instrumentation and industrialization, early access evaluation, as well as investors interested in supporting commercialization. A spin-out is envisaged.

Development Status

Experimentally validated laboratory setup (TRL 4). Label-free classification of cell-cycle states has been validated against DNA staining and confirmed by sorting and re-analysis of isolated populations; human peripheral blood T- and B-cell subsets have been classified from scatter signals alone. Development is directed towards a standardized research-use demonstrator integrating sorting.

Patent Situation

A portfolio of four patent families covers the pulse-shape readout, angle-resolved detection, and the machine-learning-based classification and sorting (US 10,337,975 B2 and EP 3 244 191 B1, priority 2016; US 11,841,314 B2 and EP 3 933 380 A1, priority 2020; EP 25193580.5, priority 2025; a fourth application covering neural-network-based classification is in preparation).