Our recent work has shown that high-throughput label-free cell classification with high accuracy can be achieved through a combination of time stretch quantitative phase imaging, microfluidics and deep learning. Such a technology holds promise for early detection of primary cancer or metastasis by finding rare diseased cells among a large population of normal cells in blood or other bodily fluids. In his February 2, 2019 talk at Photonics West, Prof. Jalali described two implementations of deep convolutional neural networks in time stretch imaging flow cytometry. In the first mode, the network operates on features extracted from cell images constructed from temporal waveforms. In the second mode, the network directly maps the raw temporal waveforms into output classes. This eliminates the image processing pipeline resulting in significantly faster runtime to enable cell sorting decisions to be made in real time.