The technique of stimulated Raman scattering microscopy has become a hot topic in recent years. The technique known as stimulated Raman scattering microscopy helps surgeons to better distinguish between cancerous and normal tissue in the patient’s brain during surgery, which may improve the safety and accuracy of such procedures. Reuben Hill, a 22-year-old doctoral student with a tumor in his brain, recently reported a first in the United Kingdom: the successful use of laser detection and a smart knife to precisely remove a brain tumor. He underwent an exceptional brain tumor removal surgery: two new technologies, laser detection and smart knife, were used during the operation. As a success story of laboratory results applied to the operating room, this groundbreaking surgery is a major reform in precision surgery. Brain tumors are more life-threatening than any other tumors, and brain tumor resection surgery is very difficult because of the interlocking and complex nerve tissues and the connection of tumor tissues with these delicate structures, sometimes it is difficult for surgeons to see the tissue structure clearly through the microscope. At the same time, removal of cancerous tissues faces great risks because the scalpel must strictly ensure that the tumor tissue is removed without damaging the surrounding normal brain tissue. Once the healthy tissue is cut, it can lead to serious side effects such as loss of speech, hearing and other functions. The new laser probe and smart knife greatly reduce these surgical risks and provide surgeons with immediate information on whether the tissue is cancerous or not. The laser probe distinguishes between cancerous and healthy tissue, and the laser can provide the surgeon with a mapping of the tumor to achieve a precise level of resection. The laser probe, which uses light reflected back from tissue by Raman spectroscopy molecules for tissue differentiation, was developed and supplied by Vancouver, Canada-based erisante Technology, which Vaqas says is the first successful application of Raman spectroscopy to human brain surgery. So what else can laser Raman spectroscopy actually do? Cancer is one of the most serious diseases that threaten human health and life. Early diagnosis and timely treatment are the most effective ways to improve the survival rate of cancer patients. As a non-invasive detection technique, laser Raman spectroscopy can provide rich information on the molecular structure and composition of substances without damage, and can be used for early diagnosis of cancer by reflecting the structural differences between cancerous and normal tissues at the molecular level. The research progress in the diagnosis of skin cancer, nasopharyngeal cancer, lung cancer, gastric cancer, colon cancer, breast cancer and prostate cancer is reviewed, and the development direction and application prospect of Raman spectroscopy in cancer diagnosis are further prospected to provide a reference basis for the early detection of cancer and the application of diagnostic techniques. In 1928, Raman, an Indian physicist, discovered the existence of Raman scattering in his research, and since then, researchers have done in-depth research on this technique, but it was not widely used because of the limitation of light source. It was not until the introduction of laser in the 1960s that the weakness of Raman spectroscopy signal was completely overcome, and Raman spectroscopy has been developed and widely used in scientific research, and laser light source has been the ideal light source for Raman spectroscopy since then. The real discovery of the SERS technique dates back to 1977, when Van Duyne and Jeanmaire systematically studied the same system as Fleischmann’s group. It was only after the systematic study of the same system as that of Fleischmann’s group that Van Duyne and Jeanmaire found that the Raman scattering signal of pyridine adsorbed on a rough Ag surface was enhanced by about 6 orders of magnitude compared to the same amount of pyridine in the solution phase. With the continuous improvement of medical detection and diagnostic techniques, the early clinical diagnosis of cancer has also made rapid progress. However, in most cases, biopsy is still required. However, in most cases, biopsies are still required for diagnosis, which are highly invasive, more damaging to patients, slower to detect, and may cause the spread of cancer cells, and require a certain level of knowledge in pathology. In order to solve the above problems, researchers have been constantly researching and trying new detection and analysis techniques to achieve rapid and non-invasive early clinical diagnosis of cancer. As a non-invasive detection technique, laser Raman spectroscopy can provide rich information on the molecular structure and composition of substances, which is often referred to as molecularfingerprints of substances and is expected to achieve nondestructive detection at the molecular level. Compared with traditional medical diagnostic methods, Raman spectroscopy has the advantages of non-destructive, non-invasive noninvasivedetection, high resolution, no reagents and high automation. Therefore, the application of Raman spectroscopy in the field of medical detection and diagnosis has gained much attention. In the process of tumor growth and development, the structure, conformation and quantity of substances in tissue cells will change significantly, and Raman spectroscopy can achieve highly sensitive and high-resolution detection of these information changes, and then reveal the differences between the structures of cancerous tissues and normal cells at the molecular level. Therefore, Raman spectroscopy is of great significance for the early diagnosis and timely treatment of cancer, thus improving the chances of survival of cancer patients. This paper reviews the application and research progress of Raman spectroscopy in the detection and diagnosis of various cancers, and further outlooks the application prospects of Raman spectroscopy in this field, expecting to provide reference information and guidance for the application of cancer early detection and diagnosis technology. 1.1 Raman spectroscopy in cancer detection and diagnosis 1.1 Raman spectroscopy for skin cancer detection There are three main types of skin cancer: basalcellcarcinoma (BCC), squamous cell carcinoma (SCC) and malignant melanoma (MM), among which Basal cell carcinoma is the most common. Basal cell carcinoma and squamous cell carcinoma are curable in almost all cases if treated promptly, while malignant melanoma is the least common but most serious skin cancer that can lead to death if not treated promptly. However, since the symptoms of these cancers are very similar, there are difficulties in the diagnostic process, and it is not feasible to remove every pigment of the patient’s tissue for biopsy, so a non-invasive and non-invasive method of detection and diagnosis needs to be sought.Nijssen et al. used a near-infrared excitation light source to obtain Raman spectra of basal cell carcinoma tissue, dermal tissue, and epithelial tissue, and used multivariate Statistical analysis and cluster analysis were used to analyze the spectra and establish a tissue classification model, which can identify cancerous tissue and its surrounding non-cancerous tissue with a sensitivity of 100% and specificity of 93%. Their study confirmed that Raman spectroscopy can accurately determine the extent of tumor resection, which provides a strong theoretical and experimental basis for the diagnosis and treatment of basal cell carcinoma. Short et al. investigated the changes of collagen around the tumor of nodular basal cell carcinoma and found that the role of nucleic acids, histones and proteins with motor proteins in the nucleus of cancer cells was different from their role in normal epidermal cells. And the Raman spectra of dermal tissues were obtained, and it was found that the intensity of Raman spectral lines increased at 940 cm-1 and decreased significantly at 1210 cm-1 and 1270 cm-1 in the dermal tissues around the cancer cells, indicating that the collagen in the tumor peripheral tissues was not only low, but also changed structurally. In order to use Fourier transform Raman spectroscopy (FT-RS) to distinguish squamous cell carcinoma from normal skin, Pereira et al. studied the Raman spectra of human skin biopsy tissues using 1064 nm as the excitation light and found that the Raman spectral line intensities at 860 cm-1 and 939 cm-1 in normal tissues were significantly higher than the corresponding spectral intensities in cancerous tissues, and were higher at 1,555 cm -Gniadecka et al. investigated the Raman spectral characteristics of melanoma and other skin lesions using near-infrared Fourier transform (NIRFT) Raman spectroscopy and found that the intensity of the protein amide I band decreased and the intensity of the lipid signature peak increased in malignant melanoma. The sensitivity and specificity of Raman spectroscopy for the diagnosis of malignant melanoma reached 85% and 99%, respectively, using a neural network approach for spectral analysis. Huang et al. successfully obtained the in vivo Raman spectroscopy of skin melanin using near-infrared Raman spectroscopy (NIR-RS), and analyzed the spectra to find that the Raman spectra of melanin at 1580 cm-1 and 1380 cm-1, respectively The Raman spectral bands with higher intensity and wider frequency band existed, which belonged to the planar vibration of aromatic ring and the stretching vibration mode of C-C, respectively. The spectral signals of melanin were obtained under in vivo conditions using Raman spectroscopy, indicating that Raman spectroscopy may become a very effective clinical test for in situ analysis and diagnosis of skin. Cheng et al. used microscopic Raman spectroscopy to analyze the conformational and chemical compositional changes in humanskinpilomatrixoma (PMX) and found that the Raman spectra of normal skin differed significantly from those of soft PMX and hard PMX tissues, especially the 1,665 cm-1 The characteristic peak attributed to amide I was shifted to 1655 cm-1, and the intensity of the characteristic peak attributed to amide III in the Raman spectra of hard PMX tissues was significantly reduced. These results all indicate that microscopic Raman spectroscopy can effectively distinguish between normal skin tissue, soft PMX tissue and hard PMX tissue. Moreover, Raman spectroscopy has high accuracy in identifying different skin lesions, especially for cancerous tissues, and has great potential application in the in situ analysis and diagnosis of malignant tumors. 1.2 Raman spectroscopy of nasopharyngeal and lung cancers Lau et al. examined the nasopharyngeal biopsy specimens by Raman spectroscopy, and the acquisition time of each spectrum was only 5 s. The analysis revealed that the Raman spectral intensity of cancerous tissues in the wave number range of 1290 cm-1 to 1320 cm-1 and 1420 cm-1 to 1470 cm-1 was larger than that of normal tissues, while the intensity of the corresponding spectral lines in the wave number range of 1530 cm-1 to 1580 cm-1 was larger than that of normal tissues. ～In the wave number range of 1530 cm-1 to 1580 cm-1, the intensity of Raman spectra in normal tissues was greater than that in cancerous tissues. They also studied the normal tissue, cancerous tissue, and squamous cell papilloma of the larynx separately using Raman spectroscopy. The peak analysis of Raman spectra showed that the sensitivity of normal tissue, cancerous tissue and squamous cell papilloma were 89%, 69% and 88%, respectively, and the specificity was 86%, 94% and 94%, respectively. The visible spectral differences were in the 850 cm-1 to 950 cm-1 and 1200 cm-1 to 1350 cm-1 spectral bands, and the relative intensity of the nucleic acid peaks increased with the progression of the lesion to malignancy. To investigate the feasibility of Raman spectroscopy for early optical detection and diagnosis of lung cancer, Huang et al. investigated the spectral information of lung cancer and normal bronchial tissues using fast diffuse near-infrared Raman spectroscopy (NIR-RS). It was shown that the Raman spectra of lung cancer and normal bronchial tissues were significantly different, and it was found that the intensity ratio of Raman spectral lines, amide I1 445/amide I1 655, was used as a judgment criterion to effectively distinguish normal lung tissues from cancerous tissues, and when amide I1 445 /amide I1 655 >1, the detected tissues were normal; amide I1445/amide I1 655 <1< span="">? when the tissue is cancerous. Yamazaki et al. constructed a new near-infrared multichannel Raman system (near-infraredmultichannelRamansystem) for the acquisition of lung tissue Raman spectra, which has high signal-to-noise ratio, avoids fluorescence It has the advantages of high signal-to-noise ratio, avoiding fluorescence interference, and short measurement time (1s). The sensitivity and specificity of the system reached 91% and 97%, respectively, when 210 lung cancer tissues and normal tissues were collected. The Raman spectra were not detected when 785 nm was used as the excitation light, while the Raman spectra with high signal-to-noise ratio were obtained when 1064 nm was used as the excitation light. Li et al. studied the fluorescence spectra and Raman spectra of serum during the development of lung cancer. The three Raman peaks belonging to β-carotene (located at 539 nm, 544 nm and 556 nm, respectively) decreased in intensity and finally disappeared. The results of this experiment suggest that the content of β-carotene gradually decreases during the deterioration of lung cancer, which can be used as a basis for diagnosing whether the lung is cancerous or not. 1.3 Detection of gastric and colon cancer Ling et al. used Fourier transform Raman (FT-Raman) spectroscopy to study 40 cases of gastric cancer and normal tissues of the stomach, and statistical processing of the spectra revealed that amide I3 240/amide I2940,amide I1660 /amide I1450,amide I1080 /amide I1450 were significantly higher in gastric cancer tissues (3240 cm -1,2940 cm-1,1660 cm-11450 cm-1,1080 cm-1 are the characteristic peak positions of OH stretching vibration of protein N-H and water, C-H stretching vibration of lipids, H-O-H variable angle vibration of protein amide I band and water, CH3 or δCH2, PO stretching vibration of nucleic acids, respectively). Therefore, these features can be used as one of the bases to discriminate whether the tissue is cancerous or not. Tang Weiyue et al. collected Raman spectra of normal and cancerous tissues in the gastric sinus, and the results showed that the 1089 cm-1 line in the Raman spectrum of cancerous tissues was significantly enhanced compared with the corresponding lines of normal tissues, and the 1459 cm-1 line was split. The extraction of this information is expected to provide a judgment basis for the detection and analysis of tumor tissues. To detect the difference between gastrointestinal cancer cells and normal cells, Yan et al. studied individual cells of gastrointestinal cancer patients using confocal micro Raman spectroscopy. The results showed that in cancer cells, the half-width of the spectral lines belonging to phenylalanine at 1002 cm-1 became narrower, the intensity of the spectral lines of leukocytes was small and there were few spectral lines, while the intensity of the spectral lines of erythrocytes was large and abundant, and there were spectral lines of pyrrole ring CN respiratory stretching vibration in the range of 1620 cm-1 to 1540 cm-1. The Raman spectra of gastric cancer cells were similar to those of normal cells, but the intensity of the spectral lines was reduced and some of them were extinguished.Huang et al. conducted a study using near-infrared Raman spectroscopy to distinguish malignant tumors from normal tissues and benign tumors by performing Raman spectroscopy on 105 colon samples, and the Raman spectra were collected under ex vivo conditions in the wave number range of 800 cm-1 to 1800 cm-1 The high resolution Raman spectrograms were collected under ex vivo conditions and analyzed to find the spectral differences between normal and cancerous tissues, and a diagnostic algorithm was established using the intensity ratio of 1002 cm-1 to 1445 cm-1 spectral line as the horizontal coordinate and the intensity ratio of 1085 cm-1 to 1445 cm-1 spectral line as the vertical coordinate, which had a sensitivity of 100% and a specificity of 96.6 Chen et al. combined the laser optical tweezers technique with Raman spectroscopy to study individual cells of epithelial carcinoma, and the collected spectra were analyzed by principal component analysis, and then logarithmic regression was performed to obtain the parametric equation that could most effectively distinguish cancer cells from normal cells. The overall sensitivity of this diagnostic model was 82.5% and the specificity was 92.5%. The study by Yan et al. found that the Raman spectral lines of intestinal cancer cells were weak and many of them disappeared, and the fluorescence intensity varied in different locations within the cancer cells. This indicates that Raman spectroscopy can provide an effective means for the early detection and diagnosis of intestinal cancer. 1.4 Raman spectroscopy in breast cancer detection Haka et al. analyzed the chemical composition of microcalcifications in benign and malignant breast lesions by Raman spectroscopy and classified them into type I calcium oxalate and type II hydroxyapatite. Type I was diagnosed as benign, and type II was both benign and malignant. The benign and malignant type II microcalcifications could be distinguished by principal component analysis of Raman spectra with a sensitivity and specificity of 88% and 93%, respectively. Haka et al. also used a linear combination model with the fitting coefficients of fat and collagen as parameters to identify 130 Raman spectrograms of normal, fibrous, and infiltrating cancer tissues and obtained a sensitivity of 94% and specificity of 96% for the Bitar et al. used Fourier transform Raman spectroscopy FT-RS to study normal breast tissues and cancerous breast tissues including different cancer subtypes. Comparing the intensity changes of the characteristic peaks in the Raman spectra of different tissues, seven different tissues could be distinguished, including normal tissue, fibrocystic tissue, in situ ductal carcinoma, in situ ductal carcinoma presenting necrotic tissue, invasive ductal carcinoma, collagen-infiltrating ductal carcinoma, and invasive lobular carcinoma. Yuanli Zhao et al. examined the Raman spectra of tissues around the periphery of surgically resected breast tumors (about 5 mm on the side of the mass) in 40 cases using microscopic confocal Raman spectroscopy. It was shown that 1440/1530 and 1082/1156 were distinguishable in the Raman spectra of tissues around breast masses with different properties, and the detection targets could be identified and classified with 1.25 and 1.03 as the boundaries, respectively. Yan Zhuan-Leung et al. studied the Raman spectra of normal breast cells and cancer cells in breast cancer patients. The Raman spectra of cancer cells were observed to be weaker overall, and the two phosphate backbone peaks 782 cm-1, 1084 cm-1 and deoxyribose-phosphate vibrational peaks 1155 cm-1 and 1262 cm-1 attributed to DNA were significantly reduced; the characteristic peaks 812 cm-1 and 979 cm-1 characterizing the A-type (DNA) conformation, 668 cm-1 disappeared, and a new peak 1175cm-1 appeared, and the spectral line of 905cm-1 was enhanced and red-shifted by 6cm-1, which indicated that there was some breakage of the phosphate backbone of DNA, which led to the loss of effective control of the division and reproduction of cancer cells. A strong class of characteristic peaks closely related to calcium sclerosis was also found in the Raman spectra of cancerous tissue cells at 960 cm-1 . These research works provide a strong experimental basis for the early detection and diagnosis of breast cancer. 1.5 Spectroscopic detection of prostate cancer Crow et al. used Raman spectroscopy to detect biopsies of BPH and malignant prostate cancer under ex vivo conditions, and the analysis revealed that the concentration of glycogen decreased and the concentration of nucleic acid increased in prostate cancer tissues compared with BPH tissues. Furthermore, a linear discriminant model was constructed using principal component analysis to identify the Raman spectra of the cancer groups at different stages, which enabled the grading of prostate cancer. Crow et al. further studied four different prostate cell lines (LNCap, PCa2b; DUI45, PC3) and established a PCA/LDA diagnostic algorithm using three principal components, PC1, PC2 and PC3. PC1 represents the increased concentrations of nucleic acids (721 cm-1 , 783 cm-1 , 1305 cm-1 , 1450 cm-1 , 1577 cm-1 ), DNA backbone (827 cm-1 , 1096 cm-1 ) and disordered proteins (1250 cm-1 , 1658 cm-1 ); PC2 represents protein α-helices (935 cm-1 , 1263 cm-1, 1657 cm-1) and phospholipids (719 cm-1, 1094 cm-1, 1125 cm-1, 1317 cm-1); PC3 represents the decreased concentrations of lipids (1090 cm-1, 1302 cm-1, 1373 cm-1), glycogen (484 cm-1) and nucleic acids (786 cm-1, 1381 cm-1 , 1576 cm-1) decreasing concentrations. When the value of PC3 is large and the value of PC2 is less than or equal to 0, the two cell lines DUI45, PC3 can be identified. And when the value of PC2 is larger than 0 and PC3 is smaller or equal to 0, two cell lines, LNCap and PCa2b, can be identified. 2.Prospects of Raman spectroscopy in tumor detection Raman spectroscopy is a non-destructive, non-invasive and high resolution detection method, which has shown its advantages in cancer diagnosis. However, its disadvantages such as weak signal and susceptibility to background fluorescence interference have limited the application of Raman spectroscopy. However, the application of various Raman spectroscopy techniques such as time-resolved Raman spectroscopy, Fourier transform infrared Raman spectroscopy and the continuous research can effectively overcome the existing limitations and enable its wider promotion and application in the biomedical field. With the development of laser optical tweezers and confocal microscopy, the diagnosis of cancerous tissues at the single-cell level has been realized, which is expected to reveal the mechanism of cancer, thus establishing a more powerful experimental basis for cancer diagnosis. The introduction of fiber-optic technology allows real-time, in situ measurement of various tissues such as skin, which can reduce the risk to patients while achieving real-time, effective diagnosis and improving the chances of survival. In conclusion, with the accumulation of sample studies, improvement of research methods and equipment, optimization of statistical models, and perfect combination of various techniques with Raman spectroscopy, Raman spectroscopy will certainly move from experimental research to clinical diagnostic applications and will be widely used in cancer research and clinical diagnosis.