Introducing Discrete Frequency Infrared Technology for High-Throughput Biofluid Screening.

Accurate early diagnosis is critical to patient survival, management and quality of life. Biofluids are key to early diagnosis due to their ease of collection and intimate involvement in human function. Large-scale mid-IR imaging of dried fluid deposits offers a high-throughput molecular analysis paradigm for the biomedical laboratory. The exciting advent of tuneable quantum cascade lasers allows for the collection of discrete frequency infrared data enabling clinically relevant timescales. By scanning targeted frequencies spectral quality, reproducibility and diagnostic potential can be maintained while significantly reducing acquisition time and processing requirements, sampling 16 serum spots with 0.6, 5.1 and 15% relative standard deviation (RSD) for 199, 14 and 9 discrete frequencies respectively. We use this reproducible methodology to show proof of concept rapid diagnostics; 40 unique dried liquid biopsies from brain, breast, lung and skin cancer patients were classified in 2.4 cumulative seconds against 10 non-cancer controls with accuracies of up to 90%.

A miniature photorefractive circuit for principal component extraction.

The optical feature extractor is a photorefractive ring oscillator that can identify the strongest spatiotemporal component of its input space. The theoretical sections discuss the design and performance limitations of the signal extractor. A simple model of the filter's nonlinear functioning enables the reader to go directly to the experimental section that describes the making of the filter and experimental results. The device, also called the auto-tuning filter, is 5 cm2 in size, has a 3 GHz processing bandwidth, and requires less than 5 mW of continuous optical power to operate.