Glioma is the most common brain neoplasm that features aggressive behavior with a dismal prognosis. Isocitrate dehydrogenase (IDH) gene mutation in glioma is an early genetic event in gliomagenesis that occurs in virtually every tumor cell and can cause profound metabolic changes. In this manuscript, we report for the first time the analysis of Raman optical signatures of IDH genotypes for human glioma using visible resonance Raman (VRR) spectroscopy. We demonstrated that VRR is a rapid, label-free, and objective method as an alternative to the existing methods for the rapid intraoperative determination of IDH mutation status with high accuracy. This study shows AI-assisted VRR has the potential to provide a new optical molecular biomarker and perform early diagnosis of glioma, which is of great importance for current guiding surgical strategies and even for targeting in situ therapies in the future.

Currently, liquid biopsy method is mainly used for tumor detection based on genomic molecular alterations in vitro. Liquid biopsy is superior to traditional tissue biopsy techniques and its diagnosis time of disease and repeated diagnosis of liquid biopsy are new breakthroughs in clinical application. Liquid biopsy method can be used to detect most human disease based on genetic biomarkers from body fluids, among which, special biomarkers in blood and cerebrospinal fluid (CSF) samples are the main research objects, and have made good achievements in preliminary clinical applications. The application of optical spectroscopy in the field of liquid biopsy has aroused great interest among researchers and demonstrated the potential of its clinical application for oncology. The aim of this study is to reveal the optical spectroscopic characteristics of the main biochemical components of CSF of brain tumor using visible resonance Raman (VRR) spectroscopy ex vivo. Tumor-associated proteins, glucose, lactate and other metabolites released to CSF can be used as markers for liquid biopsy. We studied the VRR spectra of CSF samples from 7 types of brain tumor patients. The characteristic VRR modes that were found and may be used as a combination of multiple analyte biomarkers include amyloid-ss and tau protein, excess neurotransmitters such as glutamic acid derived from the exchange with interstitial fluid (ISF), DNA, glucose, lactate, etc. for optical liquid biopsy analyses. Another interesting finding was that CSF of different types of tumors showed different images similar to the crystallization of water under the optical microscope. Considering our previous study, the current study on CSF provides another proof that the VRR system can provide a complete scan region of 200 - 4000cm(-1) as a clinical tool for non-invasive diagnosis of brain disease.

Convolutional neural network (CNN) based deep learning is used to analyze spectral data collected by visible resonance Raman (VRR) spectroscopy to distinguish human glioma tumors from healthy brain tissues using binary classification and identify the cancer grades of the glioma tumors using multi-class classification. Classification was performed using both raw spectral data and baseline-subtracted data for comparison. The classification using both datasets yielded high accuracy, with the results obtained from baseline subtracted spectra slightly better than that obtained from raw spectra. The study showed VRR combined with deep learning provides a robust molecular diagnostic tool for accurately distinguishing glioma tumors from normal tissues and glioma tumor tissues at different cancer grades. Deep learning aided VRR technique may be used for in-situ intraoperative diagnosis of brain cancer. It may help a surgeon to identify cancer margins and even cancer grades during surgery.

Native fluorescence spectra of retinoic acid (RA)-treated and untreated human breast cancer cells were measured using selective wavelengths of 300 nm and 340 nm for excitation. The spectral data of the two types of cells were analyzed using machine learning algorithms for linear unmixing and classification which yielded high accuracy. The results show that the concentrations of the native fluorophores such as tryptophan, NADH and flavins in the human malignant breast cells change when they are treated with RA. The study shows the dual-wavelength fluorescence spectroscopy aided by machine learning has potential clinical applications in drug development and chemotherapeutic studies.

Alzheimer’s disease (AD) pathogenesis is widely believed to be associated with the production and deposition of the β-amyloid peptide (Aβ) and neurofibrillary tangles (NFTs) which are composed of a highly-phosphorylated form of the microtubule-associated protein tau. Based on the above hypothesis, there are currently no sufficiently effective technologies and drugs for early detection and treatment of AD. Even the most promising new drug Lecanemab that is based on an anti-amyloid monoclonal antibody therapy, has only partially slowed down the cognitive performance of patients with mild impairment caused by Alzheimer's disease. The main symptoms of AD brain tissue lesions in patients are the deposition of β-amyloid peptide and the hyperphosphorylation of tau protein, which aggregates the microtubule structure of neurons. Therefore, Aβ deposition and hyperphosphorylation of Tau are important pathological biomarkers of Alzheimer's disease. Therefore, the main targets of research for AD prevention, detection and pharmaceuticals are still Aβ and Tau protein. The aim of this study was to detect the changes of Aβ and Tau proteins in the mouse brain tissue with AD and control samples using Visible Resonance Raman (VRR) spectroscopic technology. An attempt was made to develop criteria for the detection of early AD lesions by optical spectroscopy technology. The VRR spectra of AD, the control mouse brain tissues, and Aβ and Tau proteins were recorded and analyzed. The AD and the control mouse brain tissue samples were selected from the thalamus, frontal lobe cortex and hippocampus brain areas. VRR technology with high spatial resolution and the resonance-enhanced features of certain protein molecules is first used in this study to detect and characterize the changes of Aβ and Tau proteins in AD mouse brain model. The optical spectroscopy biomarkers of AD and Control brain tissue were identified in fingerprint and the high-wavenumber regions. The Raman spectra of the secondary structure of protein in amide (I-II-III-B-A) are detected and analyzed. The results indicate that the intensity of Amide I decreased at the 1666 cm-1 corresponding to the β-sheet structure, and the intensity of the amide III bands (1220- 1320 cm-1) increased in all AD brain tissues. It was also observed that the Raman peaks of 1448 and 980 cm-1 related to the abundance of proline, serine, and threonine at tau phosphorylation sites were significantly enhanced in the frontal lobe cortex and hippocampus of AD brain tissues. The intensity ratio biomarker of high phosphorylation in the high wavenumber range from 2898 to 2932 cm-1 increased in all AD brain tissues. Changes of protein secondary conformation and abnormally phosphorylated tau or tauopathies were observed. In summary, VRR is a sensitive tool for characterizing protein structural changes and monitoring the tau phosphorylation. It may potentially be used for early detection of AD.

The Stokes shift spectra (S3) of human cancerous and normal prostate tissues were collected label free at a selected wavelength interval of 40 nm to investigate the efficacy of the approach based on three key molecules-tryptophan, collagen, and reduced nicotinamide adenine dinucleotide (NADH)-as cancer biomarkers. S3 combines both fluorescence and absorption spectra in one scan. The S3 spectra were analyzed using machine learning (ML) algorithms, including principal component analysis (PCA), nonnegative matrix factorization (NMF), and support vector machines (SVMs). The components retrieved from the S3 spectra were considered principal biomarkers. The differences in the weights of the components between the two types of tissues were found to be significant. Sensitivity, specificity, and accuracy were calculated to evaluate the performance of SVM classification. This research demonstrates that S3 spectroscopy is effective for detecting the changes in the relative concentrations of the endogenous fluorophores in tissues due to the development of cancer label free.

Simple Summary Real-time diagnosis tools and methods are desired to aid in the intraoperative grading of glioma and tumor boundary identification to achieve safe maximal tumor removal. Raman spectroscopy is an optical method for real-time glioma detection, but few studies use fresh glioma tissue for biochemical analysis. This study is the first investigation of human glioma using a portable VRR-LRRTM Raman analyzer under quasi-clinical conditions, and reveals significant spectral differences between normal (control) and different grades of glioma. A principal component analysis–support vector machine (PCA-SVM) machine learning method was used to distinguish glioma tissues from normal tissues and different glioma grades. The accuracy in identifying glioma from normal tissue was over 80%, compared with histopathology as the gold standard. This result validates the possibility of glioma diagnosis using fresh tissue and provides instant feedback for neurosurgeons in guiding maximal safe resection, and it may support the translation of this portable tool for in vivo and real-time use in tissue biochemical analysis. Abstract There is still a lack of reliable intraoperative tools for glioma diagnosis and to guide the maximal safe resection of glioma. We report continuing work on the optical biopsy method to detect glioma grades and assess glioma boundaries intraoperatively using the VRR-LRRTM Raman analyzer, which is based on the visible resonance Raman spectroscopy (VRR) technique. A total of 2220 VRR spectra were collected during surgeries from 63 unprocessed fresh glioma tissues using the VRR-LRRTM Raman analyzer. After the VRR spectral analysis, we found differences in the native molecules in the fingerprint region and in the high-wavenumber region, and differences between normal (control) and different grades of glioma tissues. A principal component analysis–support vector machine (PCA-SVM) machine learning method was used to distinguish glioma tissues from normal tissues and different glioma grades. The accuracy in identifying glioma from normal tissue was over 80%, compared with the gold standard of histopathology reports of glioma. The VRR-LRRTM Raman analyzer may be a new label-free, real-time optical molecular pathology tool aiding in the intraoperative detection of glioma and identification of tumor boundaries, thus helping to guide maximal safe glioma removal and adjacent healthy tissue preservation.

In this study, we trained a convolutional neural network (CNN) utilizing a mix of recent CNN architectural design strategies. Our goals are to leverage these modern techniques to improve the binary classification of kidney tumor images obtained using Multi-Photon Microscopy (MPM). We demonstrate that incorporating these newer model design elements, coupled with transfer learning, image standardization, and data augmentation, leads to significantly increased classification performance over previous results. Our best model averages over 90% sensitivity, specificity, accuracy, area under the receiver operating characteristic curve (AUROC) in image-level classification across cross-validation folds, superior to the previous best in all four metrics.

We report on the use of label-free, native fluorescence (NFL) spectroscopy and machine learning (ML) algorithms to study the correlation of relative tryptophan levels with prostate cancer aggressiveness. Three extensively studied prostate cancer cell lines were used; PC3, an aggressive, androgen-resistant line, with a high tendency to metastasize in vivo, DU-145, a less aggressive cancer cell line, also androgen-resistant, and LNCaP, an androgen sensitive line, which has a low tendency to metastasize. Using an excitation of 300 nm, differences in the NFL spectral profiles from these cell lines were found to correlate with changes in the relative concentrations of tryptophan and reduced nicotinamide adenine dinucleotide (NADH). The use of ML may present a powerful tool for the assessment of the likelihood of a cancer to metastasize. This technique could aid in the decision whether to use highly aggressive adjuvant chemotherapy or radiation therapy after surgical resection of a prostate cancer.

The unique advantage of visible resonance Raman (VRR) spectroscopy using 532 nm excitation wavelength for biological samples is the resonance enhancement of vibrational modes of chemical bonds from cells and tissues. The aim of this study is specifically to reveal the VRR characteristic spectra of different organs in mice, find the molecular alterations in the development of white matter and gray matter of mouse embryos at different ages and study the VRR spectral information of the mouse embryo head using VRR technology.

There is still a lack of reliable intraoperative tools for glioma diagnosis and to guide the maximal safe resection of glioma. We report continuing work on the optical biopsy method to detect glioma grades and assess glioma boundaries intraoperatively using the VRR-LRRTM Raman analyzer, which is based on the visible resonance Raman spectroscopy (VRR) technique. A total of 2220 VRR spectra were collected during surgeries from 63 unprocessed fresh glioma tissues using the VRR-LRRTM Raman analyzer. After the VRR spectral analysis, we found differences in the native molecules in the fingerprint region and in the high-wavenumber region, and differences between normal (control) and different grades of glioma tissues. A principal component analysis-support vector machine (PCA-SVM) machine learning method was used to distinguish glioma tissues from normal tissues and different glioma grades. The accuracy in identifying glioma from normal tissue was over 80%, compared with the gold standard of histopathology reports of glioma. The VRR-LRRTM Raman analyzer may be a new label-free, real-time optical molecular pathology tool aiding in the intraoperative detection of glioma and identification of tumor boundaries, thus helping to guide maximal safe glioma removal and adjacent healthy tissue preservation.

We classified breast cancer cell lines of different metastatic potentials using visible resonance Raman spectroscopy, principal component analysis, and support vector machines. Cross validated classification accuracies of over 78% were achieved.

portable visible resonance Raman analyzer Liang Zhang, Yan Zhou, Binlin Wu, Shengjia Zhang, Ke Zhu, Chunyuan Zhang, Cheng-hui Liu, Xinguang Yu and Robert R. Alfano The Department of Neurosurgery, PLA General Hospital, Beijing, 100853, China; The Department of Neurosurgery, PLA Air Force Medical Center, Beijing, 100142, China;Physics Department and CSCU Center for Nanotechnology, Southern Connecticut State University, New Haven, CT 06515, USA;JRME Co., Ltd, Taizhou, Jiangsu, 225300, China;Institute of Physics, Chinese Academy of Sciences (CAS), Beijing, 100190, China;Institute for Ultrafast Spectroscopy and Lasers, Department of Physics, City College of New York of City University of New York, New York, NY 10031, USA.

Metastasis is the leading cause of mortalities in cancer patients due to the spreading of cancer cells to various organs. Detecting cancer and identifying its metastatic potential at the early stage is important. This may be achieved based on the quantification of the key biomolecular components within tissues and cells using recent optical spectroscopic techniques. The aim of this study was to develop a noninvasive label-free optical biopsy technique to retrieve the characteristic molecular information for detecting different metastatic potentials of prostate cancer cells. Herein we report using native fluorescence (NFL) spectroscopy along with machine learning (ML) to differentiate prostate cancer cells with different metastatic abilities. The ML algorithms including principal component analysis (PCA) and nonnegative matrix factorization (NMF) were used for dimension reduction and feature detection. The characteristic component spectra were used to identify the key biomolecules that are correlated with metastatic potentials. The relative concentrations of the molecular spectral components were retrieved and used to classify the cancer cells with different metastatic potentials. A multi-class classification was performed using support vector machines (SVMs). The NFL spectral data were collected from three prostate cancer cell lines with different levels of metastatic potentials. The key biomolecules in the prostate cancer cells were identified to be tryptophan, reduced nicotinamide adenine dinucleotide (NADH) and hypothetically lactate as well. The cancer cells with different metastatic potentials were classified with high accuracy using the relative concentrations of the key molecular components. The results suggest that the changes in the relative concentrations of these key fluorophores retrieved from NFL spectra may present potential criteria for detecting prostate cancer cells of different metastatic abilities.

Meningiomas are the most common brain neoplasms. They often show a wide distribution of locations and morphological attributes. Therefore, rendering tumor margin status is difficult during surgery. Accurate identification of tumor boundary intraoperatively is essential for total tumor removal and adjacent healthy tissue preservation. Visible resonance Raman (VRR) spectroscopy has been used for detection and diagnosis of human organ lesions since 2011. Here, we report for the first time the preliminary results on the evaluation of a VRR-LRRTM analyzer based on VRR technique to identify human brain meningioma grades and margins.

Multiphoton microscopy images of chromophobe renal cell carcinoma and renal oncocytoma were classified using a convolutional neural network inspired by techniques in recent architectures and yielded over 70% accuracy.

Early detection of prostate cancer is critical for the success of cancer therapy. It is believed that the biochemical changes that cause the optical spectra changes would appear earlier than the histological aberration. The aim of this ex vivo study was to evaluate the ability of Stokes Shift Spectra (S3) to identify human prostate cancerous tissues from the normal. Fifteen (15) pairs of with pathologically confirmed human prostate cancerous and normal tissues underwent Stokes Shift Spectra measurements with selective wavelength interval of 40 nm. The spectra were then analyzed using machine learning (ML) algorithms to classify the two types of tissues. The ML algorithms including principal component analysis (PCA) and nonnegative matrix factorization (NMF) were used for dimension reduction and feature detection. The characteristic component spectra were used to identify the key fluorophores related to carcinogenesis. The results show that these key fluorophores within tissue, e.g., tryptophan, collagen, and NADH, have different relative concentrations between cancerous and normal tissues. A multi-class classification was performed using support vector machines (SVMs). A leave-one- out cross validation was used to evaluate the performance of the classification with the gold standard histopathological results as the ground truth. The results with high sensitivity and specificity indicate that the S3 method is effective for detecting changes of fluorophore composition in human prostate tissues due to the development of cancer.