Study on Multi-Wavelength Transmission Spectral Feature Extraction Combined With Support Vector Machine for Bacteria Identification

Chun FENG; Nan-jing ZHAO; Gao-fang YIN; Ting-ting GAN; Xiao-wei CHEN; Min CHEN; Hui HUA; Jing-bo DUAN; Jian-guo LIU

doi:10.3964/j.issn.1000-0593(2021)09-2940-05

Spectroscopy and Spectral Analysis, Volume. 41, Issue 9, 2940(2021)

Study on Multi-Wavelength Transmission Spectral Feature Extraction Combined With Support Vector Machine for Bacteria Identification

Chun FENG^{1; 2; 3;}, Nan-jing ZHAO^{1; 3; *;}, Gao-fang YIN^{1; 3; *;}, Ting-ting GAN^{1; 3;}, Xiao-wei CHEN^{1; 2; 3;}, Min CHEN^{1; 2; 3;}, Hui HUA^{1; 2; 3;}, Jing-bo DUAN^{1; 3;}, and Jian-guo LIU^{1; 3;}

Author Affiliations

1. Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Hefei 230031, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(6)

Fig. 1. Multi-wavelength transmission spectra and normalized images of S.aureus at different concentrations
(a): Spectra of S.aureus at different concentrations; (b): Normalized spectra of different concentrations;(c): Normalized spectra at high concentrations; (d): Normalized spectra at low concentrations

Download full size

View in Article

Fig. 2. Normalized trend graphs of multi-wavelength transmission spectra of five low bacteria concentration samples

Download full size

View in Article

Fig. 3. Variance analysis diagram of multi-wavelength transmission spectrum of bacteria

Download full size

View in Article

Table 1. Cosine similarity between different microbial spectra of bacteria

View table

View in Article

Table 1. Cosine similarity between different microbial spectra of bacteria

1 E.coli 2 K.pneumoniae 3 S.aureus 4 S.typhi 5 P.aeruginosa
A original B normalized
	1-2	1-3	1-4	1-5	2-3	2-4	2-5	3-4	3-5	4-5
A	0.999 4	0.964 0	0.997 9	0.988 2	0.967 0	0.997 2	0.985 2	0.946 5	0.915 2	0.995 0
B	0.999 4	0.964 0	0.997 9	0.988 2	0.967 0	0.997 2	0.985 2	0.946 5	0.915 2	0.995 0

Table 2. The difference of normalized graph of multi-wavelength transmission spectra of five kinds of bacteria

View table

View in Article

Table 2. The difference of normalized graph of multi-wavelength transmission spectra of five kinds of bacteria

Bacteria	Optical density/a.u. (200 nm)	Slope 1 (200~245 mm)	Slope 2 (250~275 mm)	Slope 3 (280~300 mm)
K. Pneumoniae	0.006 5	-62.45	-15.48	-29.96
S. Aureus	0.005 1	-35.94	-14.82	-24.62
S. Typhi	0.007 4	-81.30	-20.91	-33.71
E. Coli	0.007 5	-82.67	-13.92	-36.09
P.Aeruginosa	0.008 5	-103.49	-26.21	-30.88

Table 3. The recognition rates of five kinds of bacteria by libsvm
View table
View in Article
Table 3. The recognition rates of five kinds of bacteria by libsvm
Bacteria Train
set Predict
set recognition
rate/%
K. Pneumoniae 9 9 100.0
S. Aureus 8 7 100.0
S. Typhi 10 10 100.0
E. Coli 13 13 100.0
P. Aeruginosa 9 8 100.0
- 49 47 100.0

Tools

Get Citation

Copy Citation Text

Chun FENG, Nan-jing ZHAO, Gao-fang YIN, Ting-ting GAN, Xiao-wei CHEN, Min CHEN, Hui HUA, Jing-bo DUAN, Jian-guo LIU. Study on Multi-Wavelength Transmission Spectral Feature Extraction Combined With Support Vector Machine for Bacteria Identification[J]. Spectroscopy and Spectral Analysis, 2021, 41(9): 2940

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites