Pulsed laser diode excitation for transcranial photoacoustic imaging

TL;DR

This work analyzes long-pulse PLD excitation for transcranial photoacoustic imaging to overcome skull-induced bandwidth loss. By comparing Q-switched, constant-energy PLD, and constant-power PLD sources in ex vivo skull transmission experiments, it identifies an optimal pulse duration around 400–500 ns where transcranial PA power is maximized due to energy scaling. The findings suggest PLD excitation can achieve SNRs comparable to conventional OPO/Nd:YAG sources at lower cost and footprint, enabling more portable brain PA devices and multispectral configurations. Practical relevance hinges on suitable low-frequency PA detection and adherence to MPE limits, but the approach promises a scalable, energy-efficient path for transcranial PA sensing.

Abstract

Photoacoustic (PA) imaging of deep tissue tends to employ Q-switched lasers with high pulse energy to generate high optical fluence and therefore high PA signal. Compared to Q-switched lasers, pulsed laser diodes (PLDs) typically generate low pulse energy. In PA imaging applications with strong acoustic attenuation, such as through human skull bone, the broadband PA waves generated by nanoseconds laser pulses are significantly reduced in bandwidth during their propagation to a detector. As high-frequency PA signal components are not transmitted through skull, we propose to not generate them by increasing excitation pulse duration. Because PLDs are mainly limited in their peak power output, an increase in pulse duration linearly increases pulse energy and therefore PA signal amplitude. Here we show that the optimal pulse duration for deep PA sensing through thick skull bone is far higher than in typical PA applications. Counterintuitively, this makes PLD excitation well-suited for transcranial photoacoustics. We show this in PA sensing experiments on ex vivo human skull bone.

Figures (4)

• Figure 1: Schematic of the experimental setup. A planar absorber is illuminated by one of three photoacoustic (PA) excitation sources: (1) a Q-switched laser, (2) a pulsed laser diode (PLD) system with constant energy (CE), (3) a PLD system with constant power (CP). The generated PA wave is measured using a planar FP sensor, interrogated by a PA tomography raster scanning setup similar to Zhang et al.zhang_backward-mode_2008 -- using a narrow tunable laser as input and measuring the output voltage of a fast photodiode with a data acquisition (DAQ) card. An ex vivo human frontal bone sample is placed in between the FP sensor and the PA source for acoustic transmission measurements. The shape and duration $\tau$ of the excitation pulse is varied as illustrated by the time-excitation-power curves $P_\textrm{e}(t)$ on the right.
• Figure 2: PA excitation with a Q-switched laser with pulse durations $\tau$ of 40 to 250ns. a Pulse shapes as measured with a fast photodiode. b PA waveforms $S_\mathrm{PA}$ normalized for pulse energy $E$. c Acoustics power spectra -- Fourier transforms of the PA signals $\hat{S}_\mathrm{PA}$, normalized for pulse energy $E$.
• Figure 3: Photoacoustic (PA) measurements of a thin, planar absorber using the constant energy pulsed laser diode (CE-PLD) system. a Pulse shape of the PA excitation pulses with pulse durations $\tau$. The reconstructed reference PA image is shown in the upper right corner as a maximum intensity projection along the depth axis. Single PLD bars collimated by microlenses can be distinguished. b Individual PA waveforms $S_\mathrm{PA}$ resulting from $\tau$ of 35ns and 63ns. Dashed lines are transcranial signals after passing through frontal bone. The solid line reference waveforms are shown at 10 % scale for visualization. c Acoustic power spectra for reference and transcranial signals.
• Figure 4: PA transmission measurements through cranial bone using constant power pulsed laser diode (CP-PLD) excitation of varying excitation pulse duration $\tau$, ranging from $100ns$ to $700ns$ in increments of $100ns$. a pulse shape measured using a fast photodiode. b Reference PA signals, and c their acoustic power spectra $\hat{S}_\mathrm{PA}$. d Transcranial PA signals, and e their acoustic power spectra $\hat{S}_\mathrm{PA}$. f&g Pulse energy corrections of (d&e).