Optimized sensing of sparse and small targets using lens-free holographic microscopy

Area of Science:

• Optical microscopy
• Bio-sensing technologies
• Image processing

Background:

• Lens-free holographic microscopy provides sub-micron resolution over large fields of view (>20 mm²), ideal for detecting low-concentration targets.
• Pixel super-resolution techniques enhance resolution and signal-to-noise ratio (SNR) by merging low-resolution images, but their efficacy for small targets is not well-established.

Purpose of the Study:

• To quantitatively compare the performance of different pixel super-resolution techniques for small-target sensing in lens-free holographic microscopy.
• To evaluate the impact of regularization methods on SNR and resolution for bio-sensing applications.

Main Methods:

• Experimental comparison of three regularization techniques: no regularization, cardinal-neighbor regularization, and a novel sparsity-promoting regularization.
• Utilized analytically-calculated gradients from Bayer-pattern image sensors for the sparsity-promoting method.
• Quantified SNR and resolution using micron-scale beads at surface coverages up to ~4%.

Main Results:

• Sparsity-promoting regularization demonstrated superior performance in enhancing SNR.
• An approximate 8 dB improvement in SNR was observed with sparsity-promoting regularization compared to other methods.
• This enhancement was consistent across tested imaging conditions for small targets.

Conclusions:

• Sparsity-promoting regularization is the optimal technique for improving SNR in lens-free holographic microscopy for small-target bio-sensing.
• The novel implementation using analytical gradients from Bayer sensors offers significant advantages for resolution and sensitivity.
• This advancement is critical for applications requiring high sensitivity detection of sparse biological samples.