Limits...
Near-infrared spectroscopy as a diagnostic tool for distinguishing between normal and malignant colorectal tissues.

Chen H, Lin Z, Mo L, Wu T, Tan C - Biomed Res Int (2015)

Bottom Line: The successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to seek a reduced subset of variables/wavenumbers and build a diagnostic model of LDA.The results showed that, compared to PLS-DA, SPA-LDA provided more parsimonious model using only three wavenumbers/variables (4065, 4173, and 5758 cm(-1)) to achieve the sensitivity of 84.6%, 92.3%, and 92.3% for the training, validation, and test sets, respectively, and the specificity of 100% for each subset.It indicated that the combination of NIR spectroscopy and SPA-LDA algorithm can serve as a potential tool for distinguishing between normal and malignant colorectal tissues.

View Article: PubMed Central - PubMed

Affiliation: Yibin University Hospital, Yibin, Sichuan 644000, China ; Key Lab of Process Analysis and Control of Sichuan Universities, Yibin University, Yibin, Sichuan 644000, China.

ABSTRACT
Cancer diagnosis is one of the most important tasks of biomedical research and has become the main objective of medical investigations. The present paper proposed an analytical strategy for distinguishing between normal and malignant colorectal tissues by combining the use of near-infrared (NIR) spectroscopy with chemometrics. The successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to seek a reduced subset of variables/wavenumbers and build a diagnostic model of LDA. For comparison, the partial least squares-discriminant analysis (PLS-DA) based on full-spectrum classification was also used as the reference. Principal component analysis (PCA) was used for a preliminary analysis. A total of 186 spectra from 20 patients with partial colorectal resection were collected and divided into three subsets for training, optimizing, and testing the model. The results showed that, compared to PLS-DA, SPA-LDA provided more parsimonious model using only three wavenumbers/variables (4065, 4173, and 5758 cm(-1)) to achieve the sensitivity of 84.6%, 92.3%, and 92.3% for the training, validation, and test sets, respectively, and the specificity of 100% for each subset. It indicated that the combination of NIR spectroscopy and SPA-LDA algorithm can serve as a potential tool for distinguishing between normal and malignant colorectal tissues.

Show MeSH
Validation cost (Gcost) as a function of the number of variables selected by SPA-LDA algorithm. The arrow indicates the minimum point of the cost curve, which occurs at three wavenumbers.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4309295&req=5

fig6: Validation cost (Gcost) as a function of the number of variables selected by SPA-LDA algorithm. The arrow indicates the minimum point of the cost curve, which occurs at three wavenumbers.

Mentions: The SPA-LDA modeling resulted in only three variables/wavenumbers, which correspond to the minimum point of the validation cost curve, as the arrow indicated in Figure 6. Figure 7 gave the preprocessed mean spectra of cancerous and normal tissues by 1st derivative in the range of 8000–4000 cm−1, where the solid circle markers indicated the wavenumber positions selected by SPA-LDA. As can be seen in Figure 7, the selected three wavenumbers (4065, 4173, and 5758 cm−1) are indeed related to these characteristic points such as spectral peaks and shoulders. Similarly, Figure 8 showed the prediction performance of the final SPA-LDA model on the training, validation, and test sets. There existed 4, 1, and 2 misclassified spectra for the three sets, respectively. The sensitivity was 84.6%, 92.3%, and 92.3% for the training, validation, and test set, respectively. The specificity for each subset was the same, that is, 100%. It was clear that SPA-LDA used only three variables to achieve superior performance to PLS-DA. Why only three variables lead to better model is maybe ascribed to the fact that the NIR signal strength between different channels is considerably correlated. Often, only a variable can represent the information distributed in its adjacent variables. Such a phenomenon is also in accordance with the purpose of SPA, which is to minimize collinearity among the selected variables. Moreover, SPA-LDA proves to be less sensitive to instrumental noise and more parsimonious than the other strategies.


Near-infrared spectroscopy as a diagnostic tool for distinguishing between normal and malignant colorectal tissues.

Chen H, Lin Z, Mo L, Wu T, Tan C - Biomed Res Int (2015)

Validation cost (Gcost) as a function of the number of variables selected by SPA-LDA algorithm. The arrow indicates the minimum point of the cost curve, which occurs at three wavenumbers.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4309295&req=5

fig6: Validation cost (Gcost) as a function of the number of variables selected by SPA-LDA algorithm. The arrow indicates the minimum point of the cost curve, which occurs at three wavenumbers.
Mentions: The SPA-LDA modeling resulted in only three variables/wavenumbers, which correspond to the minimum point of the validation cost curve, as the arrow indicated in Figure 6. Figure 7 gave the preprocessed mean spectra of cancerous and normal tissues by 1st derivative in the range of 8000–4000 cm−1, where the solid circle markers indicated the wavenumber positions selected by SPA-LDA. As can be seen in Figure 7, the selected three wavenumbers (4065, 4173, and 5758 cm−1) are indeed related to these characteristic points such as spectral peaks and shoulders. Similarly, Figure 8 showed the prediction performance of the final SPA-LDA model on the training, validation, and test sets. There existed 4, 1, and 2 misclassified spectra for the three sets, respectively. The sensitivity was 84.6%, 92.3%, and 92.3% for the training, validation, and test set, respectively. The specificity for each subset was the same, that is, 100%. It was clear that SPA-LDA used only three variables to achieve superior performance to PLS-DA. Why only three variables lead to better model is maybe ascribed to the fact that the NIR signal strength between different channels is considerably correlated. Often, only a variable can represent the information distributed in its adjacent variables. Such a phenomenon is also in accordance with the purpose of SPA, which is to minimize collinearity among the selected variables. Moreover, SPA-LDA proves to be less sensitive to instrumental noise and more parsimonious than the other strategies.

Bottom Line: The successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to seek a reduced subset of variables/wavenumbers and build a diagnostic model of LDA.The results showed that, compared to PLS-DA, SPA-LDA provided more parsimonious model using only three wavenumbers/variables (4065, 4173, and 5758 cm(-1)) to achieve the sensitivity of 84.6%, 92.3%, and 92.3% for the training, validation, and test sets, respectively, and the specificity of 100% for each subset.It indicated that the combination of NIR spectroscopy and SPA-LDA algorithm can serve as a potential tool for distinguishing between normal and malignant colorectal tissues.

View Article: PubMed Central - PubMed

Affiliation: Yibin University Hospital, Yibin, Sichuan 644000, China ; Key Lab of Process Analysis and Control of Sichuan Universities, Yibin University, Yibin, Sichuan 644000, China.

ABSTRACT
Cancer diagnosis is one of the most important tasks of biomedical research and has become the main objective of medical investigations. The present paper proposed an analytical strategy for distinguishing between normal and malignant colorectal tissues by combining the use of near-infrared (NIR) spectroscopy with chemometrics. The successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to seek a reduced subset of variables/wavenumbers and build a diagnostic model of LDA. For comparison, the partial least squares-discriminant analysis (PLS-DA) based on full-spectrum classification was also used as the reference. Principal component analysis (PCA) was used for a preliminary analysis. A total of 186 spectra from 20 patients with partial colorectal resection were collected and divided into three subsets for training, optimizing, and testing the model. The results showed that, compared to PLS-DA, SPA-LDA provided more parsimonious model using only three wavenumbers/variables (4065, 4173, and 5758 cm(-1)) to achieve the sensitivity of 84.6%, 92.3%, and 92.3% for the training, validation, and test sets, respectively, and the specificity of 100% for each subset. It indicated that the combination of NIR spectroscopy and SPA-LDA algorithm can serve as a potential tool for distinguishing between normal and malignant colorectal tissues.

Show MeSH