Flow cytometry is a work horse in life sciences as it is the only tool that can inspect and sort single cells with high throughput and minimal hands on time. Instruments on the market are large and expensive, resulting in long queuing times and preventing its widespread use in hematology and in clinical routine diagnosis. Obstacles that prevent cytometry miniaturisation are: 1) the costly and tendious assembly of optical components and 2) the lack of solutions for fast fluidic cell sorting. In a previous project, we have proposed an innovative high throughput approach towards cytometry. We combine lensfree - i.e. holographic - digital imaging of cells on a high speeds CMOS active optical pixel matrix integrated with a highly parallelised microfluidic backbone that steers cell towards different outlets using ultrafast thermal bubble actuation. Lensfree cell sorters can be realized in a cheap and compact platform, as all optomechanical components are replaced by nanoelectronics, advanced imaging and signal processing technology. The achievable throughput will depend on the efficient implementation of cell classification, hence this project is focussing as well on neuro-evolutionary algorithms and reservoir computing based principles to increase the accuracy of the system. We will thereby enable high throughput sorting starting from millimeters of whole blood in a few minutes with many applications in life sciences and clinical diagnostics of diseases at the single cell level.