Nanostructured porous silicon (PSi) films have been widely studied for the past two decades as optical transducers for the detection of various molecules, with advantages of simple fabrication, high internal surface, well-established surface chemistry and unique optical properties. Despite these significant advantages, the clinical implementation of label-free PSi-based biosensors has been impaired by their insufficient sensitivity, usually in the micromolar range for protein and DNA targets. In this work, we investigate the limiting factors of PSi-based optical biosensors and design methods for their improvement. As a model system, we study PSi Fabry-Pérot thin films and utilize reflective interferometric Fourier transform spectroscopy for real-time and label-free detection of different target proteins. The selectivity of the biosensors is achieved by functionalization of the porous layer with DNA aptamers, as capture probes. We investigate the advantages of these emerging synthetic capture probes in comparison to the corresponding gold-standard antibodies. We demonstrate that a similar biosensing performance, in terms of dynamic detection range, sensitivity and selectivity, is achieved when the respective capture probe is carefully immobilized onto the PSi transducer surface, considering orientation and surface density. Nevertheless, the stability and low cost of DNA aptamers in comparison to antibodies facilitate the production, shelf-life storage, and potential reusability of these aptamer-based biosensors. To decipher the limiting factors of PSi biosensors, we derive a comprehensive mathematical model, which considers all mass transport and reaction kinetics phenomena in these biosensors. We solve the model numerically and demonstrate that the model successfully captures target binding rate in these biosensors, contrary to the conventional model used in the literature. The model is used to elucidate the orders of magnitude deviations between experimental and theoretical affinities between the capture probes and the target proteins observed in these biosensors and to develop rule of thumbs for their optimization. To enhance the performance of PSi-based biosensors, we design methods for mass transfer acceleration. These include application of isotachophoresis (ITP) method for on-chip protein concentration, target mixing on top of the biosensor or simple microfluidic integration, with up to 1000-fold enhancement in sensitivity. To allow flexible study of different microfluidic designs, we integrate for the first time PSi-based biosensor in 3D-printed polyacrylate microfluidic devices by a simple bonding method and demonstrate an improved performance of the 3D-printed microfluidics, compared to the gold-standard polydimethylsiloxane (PDMS) polymer used for microfluidic fabrication. Finally, we develop a PSi-based biosensor for detection of a relevant protein cancer biomarker and present a selective target detection in a highly complex fluid of pancreatic juice. By application of the methods described above, we were able to improve the sensitivity of the biosensor to the nanomolar range. This work paves the way towards clinical application of PSi-based biosensors and their translation to point of care settings.