Machine learning has been developed dramatically and witnessed a lot of applications in various fields over the past few years. This boom originated in 2009, when a new model emerged, that is, the deep artificial neural network, which began to surpass other established mature models on some important benchmarks. Later, it was widely used in academia and industry. Ranging from image analysis to natural language processing, it fully exerted its magic and now become the state-of-the-art machine learning models. Deep neural networks have great potential in medical imaging technology, medical data analysis, medical diagnosis and other healthcare issues, and is promoted in both pre-clinical and even clinical stages. In this review, we performed an overview of some new developments and challenges in the application of machine learning to medical image analysis, with a special focus on deep learning in photoacoustic imaging.
Conventional reconstruction algorithms (e.g., delay-and-sum) used in photoacoustic imaging (PAI) provide a fast solution while many artifacts remain, especially for limited-view with ill-posed problem. In this paper, we propose a new convolutional neural network (CNN) framework Y-Net: a CNN architecture to reconstruct the initial PA pressure distribution by optimizing both raw data and beamformed images once. The network combines two encoders with one decoder path, which optimally utilizes more information from raw data and beamformed image. We compared our result with some ablation studies, and the results of the test set show better performance compared with conventional reconstruction algorithms and other deep learning method (U-Net). Both in-vitro and in-vivo experiments are used to validated our method, which still performs better than other existing methods. The proposed Y-Net architecture also has high potential in medical image reconstruction for other imaging modalities beyond PAI.
Photoacoustic tomography (PAT), a promising medical imaging method that combines optical and ultrasound techniques, has been developing for decades mostly in preclinical application. A recent trend is to utilize the economical laser source to develop a low-cost sensing and imaging system, which aims at an affordable solution in clinical application. These low-cost laser sources have different modulation modes such as pulsed modulation, continuous modulation and coded modulation to generate different profiles of PA signals in photoacoustic (PA) imaging. In this paper, we review the recent development of the photoacoustic sensing and imaging based on the economical laser sources such as laser diode (LD) and light-emitting diode (LED) in different kinds of modulation types, and discuss several representative methods to improve the performance of such imaging systems based on low-cost laser sources. Finally, some perspectives regarding the future development of portable PAT systems are discussed, followed by the conclusion.
Diabetic patients need long-term and frequent glucose monitoring to assist in insulin intake. The current finger-prick devices are painful and costly, which places noninvasive glucose sensors in high demand. In this review paper, we list several advanced electromagnetic (EM)-wave-based technologies for noninvasive glucose measurement, including infrared (IR) spectroscopy, photoacoustic (PA) spectroscopy, Raman spectroscopy, fluorescence, optical coherence tomography (OCT), Terahertz (THz) spectroscopy, and microwave sensing. The development of each method is discussed regarding the fundamental principle, system setup, and experimental results. Despite the promising achievements that have been previously reported, no established product has obtained FDA approval or survived a marketing test. The limitations of, and prospects for, these techniques are presented at the end of this review.
Nowadays, breast cancer has increasingly threatened the health of human, especially females. However, breast cancer is still hard to detect in the early stage, and the diagnostic procedure can be time-consuming with abundant expertise needed. In this paper, we explored the deep learning algorithms in emerging photoacoustic tomography (PAT) for breast cancer diagnostics. Specifically, we used a pre-processing algorithm to enhance the quality and uniformity of input breast cancer images and a transfer learning method to achieve better classification performance. Besides, by comparing the Area Under the Curve (AUC), sensitivity, and specificity of Support Vector Machine (SVM) with AlexNet and GoogLeNet, it can be concluded that the combination of deep learning and photoacoustic imaging has the potential to achieve important impact on clinical diagnostics. Finally, according to the Breast Imaging Reporting and Data System (BI-RADS) levels, we divided breast cancer images in to six grades and designed a segmentation software for identifying the six grades of breast cancer. Then, we tested based on MAMMOGRAPHYC IMAGES DATABASE FROM LAPIMO EESC/USP (Laboratory of Analysis and Processing of Medical and Dental Images) to verify the accuracy of our segmentation method, which showed a satisfactory result.