Search

Photoacoustic microscopy (PAM) is an emerging imaging method combining light and sound. However, limited by the laser's repetition rate, state-of-the-art high-speed PAM technology often sacrifices spatial sampling density (i.e., undersampling) for increased imaging speed over a large field-of-view. Deep learning (DL) methods have recently been used to improve sparsely sampled PAM images; however, these methods often require time-consuming pre-training and large training dataset with ground truth. Here, we propose the use of deep image prior (DIP) to improve the image quality of undersampled PAM images. Unlike other DL approaches, DIP requires neither pre-training nor fully-sampled ground truth, enabling its flexible and fast implementation on various imaging targets. Our results have demonstrated substantial improvement in PAM images with as few as 1.4$\%$ of the fully sampled pixels on high-speed PAM. Our approach outperforms interpolation, is competitive with pre-trained supervised DL method, and is readily translated to other high-speed, undersampling imaging modalities.

Optoacoustic imaging is a medical imaging modality that uses optical excitation and acoustic detection to generate images of tissue structures based up optical absorption within a tissue sample. This imaging modality has been widely explored as a tool for a number of clinical applications, including cancer diagnosis and wound healing tracking. Recently, the optoacoustic imaging community has published a number of reports of hand-held optoacoustic imaging devices and platforms; these hand-held configurations improve the modality’s potential for commercial clinical implementation. Here, we review recent advancements in hand-held optoacoustic imaging platforms and methods, including recent pre-clinical applications, and we present an overview of the remaining limitations in optoacoustic imaging that must be addressed to increase the translation of the modality into commercial and clinical use. Keywords: Hand-held, Optoacoustic tomography, Photoacoustic microscopy, Multispectral, Raster scanning

We propose a differential-type complementary metal-oxide-semiconductor (CMOS) sub-terahertz (THz) detector with a subthreshold preamplifier. The proposed detector improves the voltage responsivity and effective signal-to-noise ratio (SNR) using the subthreshold preamplifier, which is located between the differential detector device and main amplifier. The overall noise of the detector for the THz imaging system is reduced by the preamplifier because it diminishes the noise contribution of the main amplifier. The subthreshold preamplifier is self-biased by the output DC voltage of the detector core and has a dummy structure that cancels the DC offsets generated by the preamplifier itself. The 200 GHz detector fabricated using 0.25 μm CMOS technology includes a low drop-out regulator, current reference blocks, and an integrated antenna. A voltage responsivity of 2020 kV/W and noise equivalent power of 76 pW/√Hz are achieved using the detector at a gate bias of 0.5 V, respectively. The effective SNR at a 103 Hz chopping frequency is 70.9 dB with a 0.7 W/m² input signal power density. The dynamic range of the raster-scanned THz image is 44.59 dB.

In photoelasticity, the method of obtaining the individual values of principal stresses/normal stresses separately is referred to as stress separation. Shear difference method is one of the widely used techniques for stress separation in digital photoelasticity. Normally a simple raster scanning approach is used in shear difference method in which stress separation is carried out for all the lines within the model domain by either row-wise horizontally or column-wise vertically starting from the boundary pixels. This requires the presence of a free boundary to start the integration scheme for every row of interest, which is not always possible in most of the practical problems. In order to overcome this, in this paper, an adaptive scanning scheme is proposed so that stress separation can be carried out even if there is only one free boundary pixel available in the model. The new scanning scheme is validated using the theoretically generated data for the problem of a ring subjected to internal pressure. Later, the applicability of this scheme is demonstrated by using two other example problems. � 2008 Elsevier Ltd. All rights reserved.