Evaluating oral epithelial dysplasia classification system by near-infrared Raman spectroscopy.

Until now, the classification system of oral epithelial dysplasia is still based on the architectural and cytological changes, which relies on the observation of pathologists and is relatively subjective. The purpose of present research was to discriminate the oral dysplasia by the near-infrared Raman spectroscope, in order to evaluate the classification system. We collected Raman spectra of normal mucosa, oral squamous cell carcinoma (OSCC) and dysplasia by near-infrared Raman spectroscope. The biochemical variations between different stages were analyzed by the characteristic peaks in the subtracted mean spectra. Gaussian radial basis function support vector machines (SVM) were used to establish the diagnostic models. At the same time, principal component analysis (PCA) and linear discriminant analysis (LDA) were used to verify the results of SVM. Raman spectral differences were observed in the range between 730~1913 cm-1. Compared with normal mucosa, high contents of protein and DNA in oral dysplasia and OSCC were observed. There were no significant or gradual variation of Raman peaks among different dysplastic grades. The accuracies of comparison between mild, moderate, severe dysplasia with OSCC were 100%, 44.44%, 71.15%, which elucidated the low modeling ability of support vector machines, especially for the moderate dysplasia. The analysis by PCA-LDA could not discriminate the stages, either. Combined with support vector machines, near-infrared Raman spectroscopy could detect the biochemical variations in oral normal, OSCC and dysplastic tissues, but could not establish diagnostic model accurately. The classification system needs further improvements.

Discrimination of parotid neoplasms from the normal parotid gland by use of Raman spectroscopy and support vector machine.

Preoperative diagnosis of neoplasms in the parotid gland is essential for successful surgical treatment. The purpose of this study is to apply Raman spectroscopy in order to distinguish the spectral differences between pleomorphic adenoma and Warthin tumor from that of normal parotid gland tissues. Furthermore we establish the diagnostic model of the Raman spectra of neoplasms in parotid gland by employing support vector machine (SVM) with Gaussian radial basis function. Firstly, Raman spectra from different histopathological tissues were obtained by near-infrared Raman microscope, SVM was then employed to analyze the different spectra and establish a discriminating model. As a result, the differences of peaks in the region 800-1800 cm(-1) demonstrated the biochemical molecular alterations between different histopathological tissues. Compared with normal parotid gland tissues, the content of proteins, lipids and DNA increased in pleomorphic adenoma. The content of DNA increased but proteins and lipids decreased in Warthin tumor. SVM displayed a powerful role in the classification of three different groups. The sensitivities and specificities of discrimination between different groups reached above 95% and 99%, respectively. Raman spectroscopy combined SVM algorithm could have great potential for providing a noninvasive, effective and accurate diagnostic technology for neoplasm diagnosis in the parotid gland.

[Near infrared Raman spectral character and diagnostic value of squamous cell carcinoma of oral mucosa].

OBJECTIVE: To evaluate the value of the near infrared Raman spectroscope in diagnosing oral squamous cell carcinoma (OSCC). METHODS: Near infrared Raman spectra of ten normal mucosa, twenty OSCC and thirty oral leukoplakia (OLK) cases were collected in the research. Based on the previous researches, the information of the subtracted spectra of compared group was gained by the characteristic band in them. A Gaussian radial basis function support vector machine was used to classify spectra and establish the diagnostic models. The efficacy and validity of the algorithm were evaluated. RESULTS: By analyzing the subtracted mean spectra, the increasing peak intensity in wavenumber range of 500-2 200 cm(-1) hinted us of the high contents of DNA, protein and lipid in OSCC, which elucidate the high proliferative activity. The increasing peak intensity in the wavenumber range of 500-2 200 cm(-1) hinted us of the high contents of DNA, protein and lipid in OSCC, which elucidate the high proliferative activity, but the difference between OLK and OSCC was not as much as that between normal and OSCC. The Gaussian radial basis function support vector machine showed powerful ability in grouping and modeling of normal and OSCC, and the specificity, sensitivity and accuracy were 100%, 97.44% and 98.81% correspondingly. The algorithm showed good ability in grouping and modeling of OLK and OSCC, the specificity, sensitivity and accuracy were 95.00%, 86.36% and 96.30%. CONCLUSION: Combined with support vector machines, near infrared Raman spectroscopy could detect the biochemical variations in oral normal, OLK and OSCC, and establish diagnostic model accurately.

Local sequence information-based support vector machine to classify voltage-gated potassium channels.

In our previous work, we developed a computational tool, PreK-ClassK-ClassKv, to predict and classify potassium (K+) channels. For K+ channel prediction (PreK) and classification at family level (ClassK), this method performs well. However, it does not perform so well in classifying voltage-gated potassium (Kv) channels (ClassKv). In this paper, a new method based on the local sequence information of Kv channels is introduced to classify Kv channels. Six transmembrane domains of a Kv channel protein are used to define a protein, and the dipeptide composition technique is used to transform an amino acid sequence to a numerical sequence. A Kv channel protein is represented by a vector with 2000 elements, and a support vector machine algorithm is applied to classify Kv channels. This method shows good performance with averages of total accuracy (Acc), sensitivity (SE), specificity (SP), reliability (R) and Matthews correlation coefficient (MCC) of 98.0%, 89.9%, 100%, 0.95 and 0.94 respectively. The results indicate that the local sequence information-based method is better than the global sequence information-based method to classify Kv channels.