Importance evaluation of spectral lines in Laser-induced breakdown spectroscopy for classification of pathogenic bacteria

Wang, Qianqian; Teng, Geer; Qiao, Xiaolei; Zhao, Yu; Kong, Jinglin; Dong, Liqiang; Cui, Xutai

doi:10.1364/boe.9.005837

Cited by 27 publications

(18 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to compress the long analyzing time of large dimension data like whole spectral range data, we have proposed and validated the decreasing of Gini index of RF algorithm that could be used to extract important feature lines for classification in our previous work. [ 3 ] As a filter method, the average Gini index decreasing is defined as the important weight of each feature. According to the important weights evaluated by RF in descending order, the most important features were selected.…”

Section: Resultsmentioning

confidence: 99%

“…The LIBS experimental setup has been described in detail in previous work. [ 3 ] To facilitate the understanding of this work, we briefly mention it here. As Figure S1 shows, an external trigger mode LIBS setup was used in experiments.…”

Section: Methodsmentioning

confidence: 99%

“…Laser‐induced breakdown spectroscopy (LIBS) and Raman spectroscopy are two famous spectral techniques because of their advantages in rapid, online, and in‐site detection. [ 1,2 ] As a new elemental detection method, LIBS has been used in identification of many kinds of substance with chemometrics, for instance, biomedical samples, [ 3–5 ] forensic evidences, [ 6–9 ] explosives, [ 10 ] plants, [ 11 ] alloy, [ 12 ] ore, [ 13 ] nutrient food, [ 14–16 ] and pathogenic microorganisms. [ 17 ] However, for substances with the same elemental composition, the feature line wavelengths are always similar.…”

Section: Introductionmentioning

confidence: 99%

“…We have proposed that LIBS combined with support vector machine (SVM) can be used to identify pathogenic bacteria in clinical field. [ 3 ] However, the whole spectra have large dimensions, and this costs a long time for analysis. Meanwhile, in an SVM classification model, the kind of kernel function, the cost parameter ( c ) in classifier, and the gamma parameter ( g ) in kernel function need to be set when building the model.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Feature‐level fusion of laser‐induced breakdown spectroscopy and Raman spectroscopy for improving support vector machine in clinical bacteria identification

Teng

Wang

Cui

et al. 2021

J Raman Spectroscopy

Self Cite

View full text Add to dashboard Cite

In clinical field, the diagnosis of many diseases and their development stages depend on the detection of the corresponding bacteria. Raman spectroscopy and laser‐induced breakdown spectroscopy (LIBS) are two novel spectral diagnostic technologies for clinical bacteria identification. Both of them have been used in clinical detection combined with optimized support vector machine (SVM). In this paper, two feature‐level fusion methods (before feature selection fusion [BFSF] and after feature selection fusion [AFSF]) were proposed to improve the performance of SVM classifier and reduce the analyzing time (including the parameter tuning, model training, and testing time) simultaneously by combining data of LIBS and Raman spectroscopy. Using the most important 10 feature lines as the inputs of the optimized classifier, the analyzing time could be reduced to 1 to 2 min and the correct classification rate (CCR) achieved 95.67%. Without optimizing SVM parameters, the two proposed methods could achieve rapid and accurate classification of pathogenic bacteria further. The AFSF method showed better results with less fusion features while the BFSF method achieved higher CCR at 100% with more features. Both two methods costed around 0.2 s for analysis. These indicate that the proposed feature‐level fusion methods can improve the performance of SVM for bacteria detection. Maintaining the highest diagnostic accuracy, they can achieve the minimum analyzing time.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Feature‐level fusion of laser‐induced breakdown spectroscopy and Raman spectroscopy for improving support vector machine in clinical bacteria identification

Teng

Wang

Cui

et al. 2021

J Raman Spectroscopy

Self Cite

View full text Add to dashboard Cite

show abstract

“…In this research, we want to propose LIBS as a potential detection method for bloodstains in crime scenes, in which the high-speed of analyzing process is very important. So, the Random Forests (RF) feature selection method proposed by us before [27] was used to select the most important 20 lines for analysis to reduce the processing time. In this way, the CCR achieved at 96.0% in 20 seconds.…”

Section: The Identification Of Bloodstainsmentioning

confidence: 99%

Identification and Determination of the Bloodstains Dry Time in the Crime Scenes Using Laser-Induced Breakdown Spectroscopy

et al. 2019

Self Cite

View full text Add to dashboard Cite

The bloodstain is one kind of the most important evidences in crime scenes. Laser-induced breakdown spectroscopy (LIBS) technology was used in this paper to distinguish bloodstains from analogs and determine the time since deposited (TSD) of bloodstains. Combined with support vector machine classifier, the identification results of bloodstains from analogs achieved 96% in 20 s. Four kinds of substrates were chosen to simulate the grounds and lethal tools in crime scenes. The linear relationship between single spectral line intensity and TSD was analyzed. The Pearson linear correlation coefficient (Pearson's r) calculated on the stone bricks calibration result is higher than 0.9. The spectral line intensity ratio was proposed to improve the linear calibration relationship with TSD. Building the calibration relationship between intensity ratios and TSDs, the Pearson's r and R 2 values for results on all four substrates were improved. For stone bricks, ceramic tiles, and copper plates, the correlation coefficients were all higher than 0.95. Especially for results on stone bricks, the Pearson's r achieved 0.99024 and the R 2 value achieved 0.97571. The optimal single lines for calibration on four substrates were all related to Fe element and the optimal line intensity ratios were all ratios of an Fe line to a line of other elements.

show abstract

Applications of Laser‐Induced Breakdown Spectroscopy for Trace Detection in Explosives

Wang

Teng

2023

Laser Induced Breakdown Spectroscopy (LIBS)

View full text Add to dashboard Cite

Importance evaluation of spectral lines in Laser-induced breakdown spectroscopy for classification of pathogenic bacteria

Cited by 27 publications

References 38 publications

Feature‐level fusion of laser‐induced breakdown spectroscopy and Raman spectroscopy for improving support vector machine in clinical bacteria identification

Feature‐level fusion of laser‐induced breakdown spectroscopy and Raman spectroscopy for improving support vector machine in clinical bacteria identification

Identification and Determination of the Bloodstains Dry Time in the Crime Scenes Using Laser-Induced Breakdown Spectroscopy

Applications of Laser‐Induced Breakdown Spectroscopy for Trace Detection in Explosives

Contact Info

Product

Resources

About