Due to the ability of overcoming both the dimensionality and the collinear problems of the spectral data, partial least squares ( PLS ) is in ever increasingly used for quantitative spectrometric analysis, especially for near-infrared spectrum, mid-infrared spectrum and Raman spectrum. In this work, an improved PLS algorithm is proposed for efficient information extraction and noise reduction. The spectral variables are clustering to several subsets, and several sub-models are built for each subset. Then, the sub-models are re-weighted and ensemble to the final model. Experiments on two near-infrared datasets ( octane number prediction in gasoline and nicotine prediction in tobacco leafs ) demonstrate that the new method provides superior prediction performance and outperformed the conventional PLS algorithm, and the root mean square error of prediction ( RMSEP) is reduced by 32% and 22%, respectively.

Objective To investigate the effect of the multi-material artifact reduction (MMAR) algorithm of wide-detector CT system in reducing the beam hardening artifacts in brain CT imaging. Methods Nine tubes with various iodine concentrations (0.1-16.0 mgI/ml) were placed in a uniform phantom filled with soft-tissue equivalent material. The phantom was scanned using different combinations of the tube voltage and current as follows:80 kV/530 mA, 100 kV/295 mA, 120 kV/190 mA and 140 kV/135 mA. The scanning was performed using the GE Discovery 750 and GE Revolution CT scanners, respectively. The CT values and standard deviations of the uniform areas between tubes were measured. The artifact index (AI) was calculated by using the standard deviation value outside the tubes as background noise. The artifact index values under different kV/mA combinations with different scanners were compared. CT brain images of 36 patients (n=18 on Discovery CT and n=18 on Revolution CT) were randomly selected. CT values of normal brain tissue and dark bands areas in the posterior fossa were measured for each case. The AI was calculated for these cases as for the phantom study. Paired t test was performed for phantom data analysis, and independent t test was performed for the clinical cases data analysis. Results The average AI values with Revolution CT(4.96±1.39, 4.80±1.57, 4.56±1.45, 4.76±1.57) were smaller than those of Discovery 750 (11.90 ± 6.61, 11.17 ± 5.61, 8.85 ± 4.59, 8.77 ± 3.85) under different tube voltage settings(t=3.714, 4.186, 3.745, 4.634,P<0.001). The higher the iodine concentration difference between tube pairs was, the higher the artifact index;As for clinical data, the difference in AI values between Revolution CT(2.31 ± 0.95) and Discovery 750(3.91 ± 1.32) was found statistically significant(t=4.066,P<0.001). Conclusion The multi-material artifact reduction algorithm implemented on the wide-detector Revolution CT scanner can significantly reduce beam hardening artifacts.

Objective To investigate the effect of the multi-material artifact reduction (MMAR) algorithm of wide-detector CT system in reducing the beam hardening artifacts in brain CT imaging. Methods Nine tubes with various iodine concentrations (0.1-16.0 mgI/ml) were placed in a uniform phantom filled with soft-tissue equivalent material. The phantom was scanned using different combinations of the tube voltage and current as follows:80 kV/530 mA, 100 kV/295 mA, 120 kV/190 mA and 140 kV/135 mA. The scanning was performed using the GE Discovery 750 and GE Revolution CT scanners, respectively. The CT values and standard deviations of the uniform areas between tubes were measured. The artifact index (AI) was calculated by using the standard deviation value outside the tubes as background noise. The artifact index values under different kV/mA combinations with different scanners were compared. CT brain images of 36 patients (n=18 on Discovery CT and n=18 on Revolution CT) were randomly selected. CT values of normal brain tissue and dark bands areas in the posterior fossa were measured for each case. The AI was calculated for these cases as for the phantom study. Paired t test was performed for phantom data analysis, and independent t test was performed for the clinical cases data analysis. Results The average AI values with Revolution CT(4.96±1.39, 4.80±1.57, 4.56±1.45, 4.76±1.57) were smaller than those of Discovery 750 (11.90 ± 6.61, 11.17 ± 5.61, 8.85 ± 4.59, 8.77 ± 3.85) under different tube voltage settings(t=3.714, 4.186, 3.745, 4.634,P<0.001). The higher the iodine concentration difference between tube pairs was, the higher the artifact index;As for clinical data, the difference in AI values between Revolution CT(2.31 ± 0.95) and Discovery 750(3.91 ± 1.32) was found statistically significant(t=4.066,P<0.001). Conclusion The multi-material artifact reduction algorithm implemented on the wide-detector Revolution CT scanner can significantly reduce beam hardening artifacts.