Imaging and detection of long-lived fluorescence probes in presence of highly emissive and scattering background

Abstract

Optical biomedical imaging and diagnostics is a rapidly growing field that provides both structural and functional information with uses ranging from fundamental to practical clinical applications. Nevertheless, imaging/visualizing fluorescence objects with high spatial resolution in a highly scattering and emissive biological medium continues to be a significant challenge. A fundamental limiting factor for imaging technologies is the signal-to-background ratio (SBR). For a long time to improve the SBR, we tried to improve the brightness of fluorescence probes. Many novel fluorophores with improved brightness (almost reaching the theoretical limit), redshifted emission, highly improved photostability, and biocompatibility greatly helped advance fluorescence detection and imaging. However, autofluorescence, scattering of excitation light, and Raman scattering remain fundamental limiting problems that drastically limit detection sensitivity. Similarly, significant efforts were focused on reducing the background. High-quality sample purification eliminates the majority of autofluorescence background and in a limited confocal volume allows detection to reach the ultimate sensitivity to a single molecule. However, detection and imaging in physiological conditions does not allow for any sample (cells or tissue) purification, forcing us to face a fundamental limitation. A significant improvement in limiting background can be achieved when fluorophores with a long fluorescence lifetime are used, and time-gated detection is applied. However, all long-lived fluorophores present low brightness, limiting the potential improvement. We recently proposed to utilize multipulse excitation (burst of pulses) to enhance the relative signal of long-lived fluorophores and significantly improve the SBR. Herein, we present results obtained with multipulse excitation and compare them with standard single-pulse excitation. Subtraction of images obtained with a single pulse from those obtained with pulse burst (differential image) highly limits background and instrumental noise resulting in more specific/sensitive detection and allows to achieve greater imaging depth in highly scattering media, including skin and tissue.

Keywords: DNA intercalator; Multipulsing; bioimaging; biomedical; fluorescence spectroscopy; time-resolved imaging.

Figure 1.

Schematic of excitation and emission of a fluorescent object in tissue. Excitation beam (blue), ballistic excitation photons reaching the sample, scattered photons reaching the sample, specular reflectance, and emission (red) – direct and diffused (scattered). (A color version of this figure is available in the online journal.)

Figure 2.

Simulated normalized emission spectra of background and probe (A) and expected steady-state emission of background, probe, and their composition (B). (A color version of this figure is available in the online journal.)

Figure 3.

Normalized intensity decays for probe (solid blue line), background (solid black line), and cumulative intensity decay (dashed red line). (A color version of this figure is available in the online journal.)

Figure 4.

Emission intensity decay in a standard single-pulse excitation where the measurement is triggered by the excitation pulse (top). The repetition rate of 500 kHz results in a temporal spacing between pulses of 2000 ns (1900 ns is indicated as a reference). Intensity decay trace when the excitation is achieved with a burst of four identical pulses with a pulse-to-pulse separation of 25 ns (pulse-to-pulse separation for 40 MHz laser repetition rate). The additional three pulses in the burst are generated before the first one. As a result, the single pulse is always aligned with the last pulse in the burst. (A color version of this figure is available in the online journal.)

Figure 5.

Expected signal from background (black solid line), probe (red solid line), and their composition (dashed blue line) with four pulses. The collection is initiated with the last pulse in the burst and the signal is collected for a time shorter than 1900 ns but sufficient for both background and probe to decay. (A color version of this figure is available in the online journal.)

Figure 6.

Differential spectrum obtained by subtracting the overall/total spectrum detected with a single pulse from the total spectrum obtained with a burst of four pulses (red solid line) and expected emission signal from the single-pulse excitation (black solid line). (A color version of this figure is available in the online journal.)

Figure 7.

Normalized differential spectrum obtained with repetition rate of 40 MHz (blue dots), normalized differential spectrum obtained with repetition rate of 80 MHz (black dashed line), and normalized emission spectrum of the probe (red solid line). (A color version of this figure is available in the online journal.)

Figure 8.

Normalized steady-state emission spectra of Rh and Ru in solution (A), and emissions for different mixes containing a different relative amount of background (Rh) and probe (Ru) as measured with single-pulse excitation (B). (A color version of this figure is available in the online journal.)

Figure 9.

Emission spectra for a 50:50 (A), 90:10 (B), and 98:2 (C) signal contribution solution of Rh and Ru as measured with one, three, and eight pulses bursts. The collection of photons was initiated with the last pulse, and the integration time was shorter than the arrival time of a subsequent pulse burst. (A color version of this figure is available in the online journal.)

Figure 10.

Differential spectra for the three Rh and Ru solutions (50:50, 90:10, and 98:2) obtained by subtracting the signal with one pulse from the signals with eight and three pulses. (A color version of this figure is available in the online journal.)

Figure 11.

Normalized differential spectra for the three Rh and Ru solutions (50:50, 90:10, and 98:2) obtained by subtracting the signal with one pulse from the signals with eight and three pulses. (A color version of this figure is available in the online journal.)

Figure 12.

Imaging setup of the two films of Rh and Ru placed in a Petri dish (A), photograph of the films (B), image captured with a single pulse (C), image obtained with eight pulses (D), and differential image (E). The scale bar is 5 mm. (A color version of this figure is available in the online journal.)

Figure 13.

Imaging setup of a large piece of Rh film on which we positioned Ru strips. Photograph of the sample (A), image obtained with one pulse (B), image obtained with eight pulses (C), and differential image (D). The scale bar is 2 mm. (A color version of this figure is available in the online journal.)

Figure 14.

Images obtained from samples submerged in different amounts of intralipid solution, 1 mL (1 mm) (A), 2 mL (2 mm) (B), and 3 mL (3 mm) (C). The inset shows a real photograph of the sample for each amount of intralipid. A 3D intensity-based plot is also reported. The scale bar is 2 mm. (A color version of this figure is available in the online journal.)

Figure 15.

Sample containing the two strips of laminated Rh and Ru PVA films implanted under the skin of a chicken drumstick (A), image obtained with one pulse (B), image obtained with eight pulses (C), and differential image (D). The images were obtained using a 375-nm laser as excitation source. The scale bar is 2 mm. (A color version of this figure is available in the online journal.)