Deterministic Lateral Displacement (DLD) devices are widely used in microfluidics for label-free, size-based separation of particles and cells, with particular promise in isolating circulating tumor cells (CTCs) for early cancer diagnostics. This study focuses on the optimization of DLD design parameters, such as row shift fraction, post size, and gap distance, to enhance the selective isolation of lung cancer cells based on their physical properties. To overcome the challenges of rare CTC detection and reduce reliance on computationally intensive simulations, machine learning models including gradient boosting, k-nearest neighbors, random forest, and multilayer perceptron (MLP) regressors are employed. Trained on a large, numerically validated dataset, these models predict particle trajectories and identify optimal device configurations, enabling high-throughput and cost-effective DLD design. Beyond trajectory prediction, the models aid in isolating critical design variables, offering a systematic, data-driven framework for automated DLD optimization. This integrative approach advances the development of scalable and precise microfluidic systems for cancer diagnostics, contributing to the broader goals of early detection and personalized medicine.

Size-based separation of bioparticles/cells is crucial to a variety of biomedical processing steps for applications such as exosomes and DNA isolation. Design and improvement of such microfluidic devices is a challenge to best answer the demand for producing homogeneous end-result for study and use. Deterministic lateral displacement (DLD) exploits a similar principle that has drawn extensive attention over years. However, the lack of predictive understanding of the particle trajectory and its induced mode makes designing a DLD device an iterative procedure. Therefore, this paper investigates a fast versatile design automation platform to address this issue. To do so, convolutional and artificial neural networks were employed to learn velocity fields and critical diameters of a wide range of DLD configurations. Later, these networks were combined with a multi-objective evolutionary algorithm to construct the automation tool. After ensuring the accuracy of the neural networks, the developed tool was tested for 12 critical conditions. Reaching the imposed conditions, the automation components performed reliably with errors of less than 4%. Moreover, this tool is generalizable to other field-based problems and since the neural network is an integral part of this method, it enables transfer learning for similar physics. All the codes generated and used in this study alongside the pre-trained neural network models are available on https://github.com/HoseynAAmiri/DLDNN.