Prostate cancer identification via photoacoustic spectroscopy and machine learning

Assessing Breast Cancer through Tumor Microenvironment Mapping of Collagen and Other Biomolecule Spectral Fingerprints─A Review

Breast cancer is a major challenge in the field of oncology, with around 2.3 million cases and around 670,000 deaths globally based on the GLOBOCAN 2022 data. Despite having advanced technologies, breast cancer remains the major type of cancer among women. This review highlights various collagen signatures and the role of different collagen types in breast tumor development, progression, and metastasis, along with the use of photoacoustic spectroscopy to offer insights into future cancer diagnostic applications without the need for surgery or other invasive techniques. Through mapping of the tumor microenvironment and spotlighting key components and their absorption wavelengths, we emphasize the need for extensive preclinical and clinical investigations.

Identification and Localization of Indolent and Aggressive Prostate Cancers Using Multilevel Bi-LSTM

Identifying indolent and aggressive prostate cancers is a critical problem for optimal treatment. The existing approaches of prostate cancer detection are facing challenges as the techniques rely on ground truth labels with limited accuracy, and histological similarity, and do not consider the disease pathology characteristics, and indefinite differences in appearance between the cancerous and healthy tissue lead to many false positive and false negative interpretations. Hence, this research introduces a comprehensive framework designed to achieve accurate identification and localization of prostate cancers, irrespective of their aggressiveness. This is accomplished through the utilization of a sophisticated multilevel bidirectional long short-term memory (Bi-LSTM) model. The pre-processed images are subjected to multilevel feature map-based U-Net segmentation, bolstered by ResNet-101 and a channel-based attention module that improves the performance. Subsequently, segmented images undergo feature extraction, encompassing various feature types, including statistical features, a global hybrid-based feature map, and a ResNet-101 feature map that enhances the detection accuracy. The extracted features are fed to the multilevel Bi-LSTM model, further optimized through channel and spatial attention mechanisms that offer the effective localization and recognition of complex structures of cancer. Further, the framework represents a promising approach for enhancing the diagnosis and localization of prostate cancers, encompassing both indolent and aggressive cases. Rigorous testing on a distinct dataset demonstrates the model's effectiveness, with performance evaluated through key metrics which are reported as 96.72%, 96.17%, and 96.17% for accuracy, sensitivity, and specificity respectively utilizing the dataset 1. For dataset 2, the model achieves the accuracy, sensitivity, and specificity values of 94.41%, 93.10%, and 94.96% respectively. These results surpass the efficiency of alternative methods.

Photoacoustic imaging in prostate cancer: A new paradigm for diagnosis and management

The global health issue of prostate cancer (PCa) requires better diagnosis and treatment. Photoacoustic imaging (PAI) may change PCa management. This review examines PAI's principles, diagnostic role, and therapeutic guidance. PAI uses optical light excitation and ultrasonic detection for high-resolution functional and molecular imaging. PAI uses endogenous and exogenous contrast agents to distinguish cancerous and benign prostate tissues with greater sensitivity and specificity than PSA testing and TRUS-guided biopsy. In addition to diagnosing, PAI can guide and monitor PCa therapy. Its real-time imaging allows precise biopsies and brachytherapy seed placement. Photoacoustic temperature imaging allows non-invasive monitoring of thermal therapies like cryotherapy, improving treatment precision and success. Transurethral illumination probes, innovative contrast agents, integration with other imaging modalities, and machine learning analysis are being developed to overcome depth and data complexity restrictions. PAI could become an essential tool for PCa diagnosis and therapeutic guidance as the field advances.

Hepatoprotective Effect Assessment of C-Phycocyanin on Hepatocellular Carcinoma Rat Model by Using Photoacoustic Spectroscopy

Hepatocellular carcinoma (HCC) is the most common primary neoplasia of the liver with elevated mortality. Experimental treatment with antioxidants has a beneficial effect on the experimental models of HCC. Arthrospira maxima (spirulina) and its phycocyanin have antitumoral action on different tumoral cells. However, it is unknown whether phycocyanin is the responsible molecule for the antitumoral effect on HCC. Photoacoustic spectroscopy (PAS) stands out among other spectroscopy techniques for its versatility of samples analyzed. This technique makes it possible to obtain the optical absorption spectrum of solid or liquid, dark or transparent samples. Previous studies report that assessing liver damage in rats produced by the modified resistant hepatocyte model (MRHM) is possible by analyzing their blood optical absorption spectrum. This study aimed to investigate, by PAS, the effect of phycocyanin obtained from spirulina on hepatic dysfunction. The optical absorption spectra analysis of the rat blood indicates the damage level induced by the MRHM group, being in concordance with the carried out biological conventional studies results, indicating an increase in the activity of hepatic enzymes, oxidative stress, Bax/Bcl2 ratio, cdk2, and AKT2 expression results, with a reduction in p53 expression. Also, PAS results suggest that phycocyanin decreases induced damage, due to the prevention of the Bax, AKT2, and p53 altered expression and the tumor progression in a HCC rat model.

Automatic detection of ischemic necrotic sites in small intestinal tissue using hyperspectral imaging and transfer learning

Acquiring large amounts of hyperspectral data of small intestinal tissue with real labels in the clinic is difficult, and the data shows inter-patient variability. Building an automatic identification model using a small dataset presents a crucial challenge in obtaining a strong generalization of the model. This study aimed to explore the performance of hyperspectral imaging and transfer learning techniques in the automatic identification of normal and ischemic necrotic sites in small intestinal tissue. Hyperspectral data of small intestinal tissues were collected from eight white rabbit samples. The transfer component analysis (TCA) method was performed to transfer learning on hyperspectral data between different samples and the variability of data distribution between samples was reduced. The results showed that the TCA transfer learning method improved the accuracy of the classification model with less training data. This study provided a reliable method for single-sample modelling to detect necrotic sites in small intestinal tissue .

Identification of different types of tumors based on photoacoustic spectral analysis: preclinical feasibility studies on skin tumors

Significance

Collagen and lipid are important components of tumor microenvironments (TME) and participates in tumor development and invasion. It has been reported that collagen and lipid can be used as a hallmark to diagnosis and differentiate tumors.

Aim

We aim to introduce photoacoustic spectral analysis (PASA) method that can provide both the content and structure distribution of endogenous chromophores in biological tissues to characterize the tumor-related features for identifying different types of tumors.

Approach

Ex vivo human tissues with suspected squamous cell carcinoma (SCC), suspected basal cell carcinoma (BCC), and normal tissue were used in this study. The relative lipid and collagen contents in the TME were assessed based on the PASA parameters and compared with histology. Support vector machine (SVM), one of the simplest machine learning tools, was applied for automatic skin cancer type detection.

Results

The PASA results showed that the lipid and collagen levels of the tumors were significantly lower than those of the normal tissue, and there was a statistical difference between SCC and BCC (p < 0.05 ), consistent with the histopathological results. The SVM-based categorization achieved diagnostic accuracies of 91.7% (normal), 93.3% (SCC), and 91.7% (BCC).

Conclusions

We verified the potential use of collagen and lipid in the TME as biomarkers of tumor diversity and achieved accurate tumor classification based on the collagen and lipid content using PASA. The proposed method provides a new way to diagnose tumors.

Photoacoustic imaging with limited sampling: a review of machine learning approaches

Photoacoustic imaging combines high optical absorption contrast and deep acoustic penetration, and can reveal structural, molecular, and functional information about biological tissue non-invasively. Due to practical restrictions, photoacoustic imaging systems often face various challenges, such as complex system configuration, long imaging time, and/or less-than-ideal image quality, which collectively hinder their clinical application. Machine learning has been applied to improve photoacoustic imaging and mitigate the otherwise strict requirements in system setup and data acquisition. In contrast to the previous reviews of learned methods in photoacoustic computed tomography (PACT), this review focuses on the application of machine learning approaches to address the limited spatial sampling problems in photoacoustic imaging, specifically the limited view and undersampling issues. We summarize the relevant PACT works based on their training data, workflow, and model architecture. Notably, we also introduce the recent limited sampling works on the other major implementation of photoacoustic imaging, i.e., photoacoustic microscopy (PAM). With machine learning-based processing, photoacoustic imaging can achieve improved image quality with modest spatial sampling, presenting great potential for low-cost and user-friendly clinical applications.

Molecular breast cancer subtype identification using photoacoustic spectral analysis and machine learning at the biomacromolecular level

Breast cancer threatens the health of women worldwide, and its molecular subtypes largely determine the therapy and prognosis of patients. However, an uncomplicated and accurate method to identify subtypes is currently lacking. This study utilized photoacoustic spectral analysis (PASA) based on the partial least squares discriminant algorithm (PLS-DA) to identify molecular breast cancer subtypes at the biomacromolecular level in vivo. The area of power spectrum density (APSD) was extracted to semi-quantify the biomacromolecule content. The feature wavelengths were obtained via the variable importance in projection (VIP) score and the selectivity ratio (Sratio), to identify the biomarkers. The PASA achieved an accuracy of 84%. Most of the feature wavelengths fell into the collagen-dominated absorption waveband, which was consistent with the histopathological results. This paper proposes a successful method for identifying molecular breast cancer subtypes and proves that collagen can be treated as a biomarker for molecular breast cancer subtyping.

Recent Advances in Contrast-Enhanced Photoacoustic Imaging: Overcoming the Physical and Practical Challenges

For decades now, photoacoustic imaging (PAI) has been investigated to realize its potential as a niche biomedical imaging modality. Despite its highly desirable optical contrast and ultrasonic spatiotemporal resolution, PAI is challenged by such physical limitations as a low signal-to-noise ratio (SNR), diminished image contrast due to strong optical attenuation, and a lower-bound on spatial resolution in deep tissue. In addition, contrast-enhanced PAI has faced practical limitations such as insufficient cell-specific targeting due to low delivery efficiency and difficulties in developing clinically translatable agents. Identifying these limitations is essential to the continuing expansion of the field, and substantial advances in developing contrast-enhancing agents, complemented by high-performance image acquisition systems, have synergistically dealt with the challenges of conventional PAI. This review covers the past four years of research on pushing the physical and practical challenges of PAI in terms of SNR/contrast, spatial resolution, targeted delivery, and clinical application. Promising strategies for dealing with each challenge are reviewed in detail, and future research directions for next generation contrast-enhanced PAI are discussed.

Biomedical Application of Photoacoustics: A Plethora of Opportunities

The photoacoustic (PA) technique is a non-invasive, non-ionizing hybrid technique that exploits laser irradiation for sample excitation and acquires an ultrasound signal generated due to thermoelastic expansion of the sample. Being a hybrid technique, PA possesses the inherent advantages of conventional optical (high resolution) and ultrasonic (high depth of penetration in biological tissue) techniques and eliminates some of the major limitations of these conventional techniques. Hence, PA has been employed for different biomedical applications. In this review, we first discuss the basic physics of PA. Then, we discuss different aspects of PA techniques, which includes PA imaging and also PA frequency spectral analysis. The theory of PA signal generation, detection and analysis is also detailed in this work. Later, we also discuss the major biomedical application area of PA technique.

Two-step proximal gradient descent algorithm for photoacoustic signal unmixing

Photoacoustic microscopy uses multiple wavelengths to measure concentrations of different absorbers. The speed of sound limits the shortest wavelength switching time to sub-microseconds, which is a bottleneck for high-speed broad-spectrum imaging. Via computational separation of overlapped signals, we can break the sound-speed limit on the wavelength switching time. This paper presents a new signal unmixing algorithm named two-step proximal gradient descent. It is advantageous in separating multiple wavelengths with long overlapping and high noise. In the simulation, we can unmix up to nine overlapped signals and successfully separate three overlapped signals with 12-ns delay and 15.9-dB signal-to-noise ratio. We apply this technique to separate three-wavelength photoacoustic images in microvessels. In vivo results show that the algorithm can successfully unmix overlapped multi-wavelength photoacoustic signals, and the unmixed data can improve accuracy in oxygen saturation imaging.

Myocardial infarct border demarcation by dual-wavelength photoacoustic spectral analysis

Myocardial infarction (MI) is a major cause of morbidity and mortality worldwide. Modern therapeutic strategies targeting the infarct border area have been shown to benefit overall cardiac function after MI. However, there is no non-invasive diagnostic technique to precisely demarcate the MI boundary till to now. In this study, the feasibility of demarcating the MI border using dual-wavelength photoacoustic spectral analysis (DWPASA) was investigated. To quantify specific molecular characteristics before and after MI, "the ratio of the areas of the power spectral densities (R APSD)" was computed from the DWPASA results. Compared to the normal tissue, MI tissue was shown to contain more collagen, resulting in higher R APSD values (p < 0.001). Cross-sectional MI lengths and the MI area border demarcated in two dimensions by DWPASA were in substantial agreement with Masson staining (ICC = 0.76, p < 0.001, IoU = 0.72). R APSD has been proved that can be used as an indicator of disease evolution to distinguish normal and pathological tissues. These findings indicate that the DWPASA method may offer a new diagnostic solution for determining MI borders.

Photoacoustic characterization of bone physico-chemical information

Osteoporosis usually alters the chemical composition and physical microstructure of bone. Currently, most clinical techniques for bone assessment are focused on the either bone microstructure or bone mineral density (BMD). In this study, a novel multi-wavelength photoacoustic time-frequency spectral analysis (MWPA-TFSA) method was introduced based on the optical absorption spectra and photoacoustic effects of biological macromolecules, which evaluates changes in bone chemical composition and microstructure. The results demonstrated that osteoporotic bones had decreased BMD, more lipids, and wider trabecular separation filled with larger marrow clusters, which were consistent with multiple gold-standard results, suggesting that the MWPA-TFSA method has the potential to provide a thorough bone physico-chemical information evaluation noninvasively and nonradiatively.

Study on a photoacoustic spectroscopy trichloromethane gas detection method based on an arched photoacoustic cavity

As an important component in photoacoustic spectroscopy gas detection systems, the performance of the photoacoustic cavity directly affects the sensitivity and resolution of the system. Based on a study of photoacoustic cavity performance, a new type of arched photoacoustic cavity is proposed. Finite element simulation software is used for modeling. By comparing the influences of the position and radius of the central sphere, the length and radius of the resonant cavity, and the radius of the buffer chamber on the performance of the photoacoustic cavity, the optimal structural size of the arched photoacoustic cavity is determined. Compared to a traditional cylindrical photoacoustic cavity with the same size, and considering the thermal viscous acoustic loss, a thermal-acoustic coupling multiphysical field simulation of the two models is carried out. The acoustic pressure signal of the arched photoacoustic cavity is 6 times that of the cylindrical photoacoustic cavity, the resonant frequency increases by 300 Hz, and the quality factor is 2.6 times that of the cylindrical photoacoustic cavity. The performance of the arched photoacoustic cavity is significantly improved. A photoacoustic spectroscopy system for the detection of chloroform gas (CHCl₃) is built based on an arched photoacoustic cavity. Detection experiments are carried out with different concentrations of chloroform. At room temperature (25 °C) and atmospheric pressure, the linear coefficient R₂ is 0.9975, and the detection sensitivity is 0.28 ppm. The system has great practical value for the detection of chloroform gas in industrial and agricultural applications.

Quick identification of prostate cancer by wavelet transform-based photoacoustic power spectrum analysis

Pathology is currently the gold standard for grading prostate cancer (PCa). However, pathology takes considerable time to provide a final result and is significantly dependent on subjective judgment. In this study, wavelet transform-based photoacoustic power spectrum analysis (WT-PASA) was used for grading PCa with different Gleason scores (GSs). The tumor region was accurately identified via wavelet transform time-frequency analysis. Then, a linear fitting was conducted on the photoacoustic power spectrum curve of the tumor region to obtain the quantified spectral parameter slope. The results showed that high GSs have small glandular cavity structures and higher heterogeneity, and consequently, the slopes at both 1210 nm and 1310 nm were high (p < 0.01). The classification accuracy of the PA time frequency spectrum (PA-TFS) of tumor region using ResNet-18 was 89% at 1210 nm and 92.7% at 1310 nm. Further, the testing time was less than 7 mins. The results demonstrated that identification of PCa can be rapidly and objectively realized using WT-PASA.

High-Throughput Recognition of Tumor Cells Using Label-Free Elemental Characteristics Based on Interpretable Deep Learning

With cancer seriously hampering the increasing life expectancy of people, developing an instant diagnostic method has become an urgent objective. In this work, we developed a label-free laser-induced breakdown spectroscopy (LIBS) method for high-throughput recognition of tumor cells. LIBS spectra were straightly collected from cells dropped on a silicon substrate and built into a deep learning model for simultaneous classification of various cancers. To interpret the result of the deep learning algorithm, gradient-weighted class activation mapping was utilized to a one-dimensional convolution neural network (1D-CNN), and the saliency maps thus obtained amplified the differences between the spectra of cell lines. Overall results showed that the 1D-CNN algorithms achieved a mean sensitivity of 94.00%, a mean specificity of 98.47%, and a mean accuracy of 97.56%. Thus, the proposed method performed satisfactorily and is seen as an interpretable classification process for cancer cell lines.