This work investigates a new technical approach in the field of flow cytometry. The fluorescent light of passing cells in a microfluidic channel is spatially modulated by a slit mask in between the channel and the detector yielding a non-periodic modulation signal for each traversing cell. Aside from a single detector in conjunction with a slit mask, it is demonstrated that pixel array detectors can also be used for implementing a spatial modulation scheme. Within the Poiseuille flow profile of the channel, cells migrate to specific equilibrium positions of their velocities which mainly depend on their size and morphology. This behavior is well known as the Segré-Silberberg effect and is a key technique of the measuring principle discussed within the framework of this thesis. For optical detection, cells are either specifically stained by fluorescent markers or tagged by fluorophore-conjugated antibodies attached to antigens at the cell surface. Due to their morphology, size and biological properties, fluorescent signals from different cell populations are primarily distinguished by their fluorescence intensity. Moreover, the velocity information of individual cells is encoded in the duration of the detected modulation signals. This parameter as well as the intensity information are recovered from raw data by appropriate digital signal processing methods. According to these parameters, different cell types can then be distinguished. Compared to conventional flow cytometry making use of fluorescent light and stray light for cell discrimination, the approach discussed here is fundamentally new and promises certain advantages. On the one hand, a larger detecting zone allows for higher mechanical robustness and less susceptibility to harsh environmental influences such as strong temperature fluctuations. On the other hand, our technique can be optimized for specific applications and relinquishes expensive optical elements. Certain methods for physical realization and implementation are evaluated and efficient digital signal processing methods are developed. A main subject of this thesis is the design of dedicated pulse compression filters enabling access to the detection parameters with maximal precision and maximal signal-to-noise-ratio (SNR) gain. For broad velocity distributions of cells, the 'balanced' filter has shown to be a suitable filter type which is discussed in detail in this work. Narrow velocity distributions are better processed by the so called 'sidelobe optimized' filter which shows much better signal dynamics and lower signal artifacts causing false detections. The underlying signal theory is mainly adapted from communications engineering and the field of RADAR technology and is discussed throughout this work. Focusing on binary modulation sequences, existing codes are evaluated for the principle of spatially modulated emission. Moreover, new criteria for performance rating are developed and new, better suited sequences are identified for our application. In the experimental part of the work, the functionality of the theoretically derived filters is demonstrated in a real application. Besides a benchmark of the SNR gain of different filter types, the detection of breast cancer cells (MCF-7) is demonstrated. Additionally, the principle is realized the first time with a pixel array detector allowing for an easier and much more flexible signal processing. Further application-specific designs aiming for miniaturization towards point-of-care diagnostics could highly benefit from this detection scheme. Examples are detector units for legionella in fresh water, CD-4 count (HIV surveillance) or detectors for circulating tumor cells in blood. | English |