Search

This paper focuses on the adverse impact of outlier points and the ambiguity of candidate localizations in 3D object detection in terms of point cloud dataset. First, outlier points can disperse real feature extracting and mislead object detection, we propose an outlier weakening strategy. The neighborhood points of each point in the point set can be established via multi-directional search algorithm, and the correlations among points in the neighborhood are figured out via self-attention mechanism, then each point representation can be enhanced with the key information from its neighborhood, thus the negative impact of outlier points will be weakened due to obtaining real knowledge of object from neighborhood context. Second, multiple proposed boxes for object localization usually containing the same sampling points, this causes vagueness in differing them from each other and leads to incorrect object positioning. This paper proposes a voxel coding strategy with adaptive pooling, the candidate boxes are divided into voxels, and each voxel is further divided into multiple columns, then they are weighted and aggregated according to the importance of each column, thus can pop out the most confident spatial voxel encodings as reliable object localization nominees. This algorithm achieves an average accuracy of 82.98% and 93.2% on the KITTI dataset Car category and ModelNet40 dataset, respectively. [ABSTRACT FROM AUTHOR]

Fipronil, an insecticide commonly used to rid livestock of fleas, lice and ticks, is banned from use in the food industry in the U.S. and European Union. The incident of poultry eggs tainted with fipronil in 2017 caused recall of millions of eggs affecting more than 40 countries. Since commercial application of fipronil is as a spray, its most immediate and most rapid detection is on surfaces. A partial least squares regression (PLSR) model was developed with infrared (IR) data (750–2000 cm−1). The relationship between the peak intensity and the fipronil concentration is linear. The method finds a correlation coefficient R2 = 0.98 in the calibration set and R2 = 0.96 in the validation set. The IR spectral assignments were made for 18 vibrational modes (750–2500 cm−1). Eighteen Raman vibrational modes in the pure fipronil have also been found (500–2500 cm−1). The molecular sites in vibrational modes for C–C and C–N stretching in both IR and Raman spectra (1250–1750 cm−1) are similar, yet the pattern of their relative intensities in both are different. Differences in the relative spectral intensity at wavenumbers 1319, 1423, 1506, and 1634 cm−1 could also be less affected by potential interferences from matrix and background signals. The dual modality IR and Raman measurements can provide a unique and highly specific fingerprint of fipronil identity. [ABSTRACT FROM AUTHOR]

Food safety and foodborne diseases are significant global public health concerns. Meat and poultry carcasses can be contaminated by pathogens like E. coli and salmonella, by contact with animal fecal matter and ingesta during slaughter and processing. Since fecal matter and ingesta can host these pathogens, detection, and excision of contaminated regions on meat surfaces is crucial. Fluorescence imaging has proven its potential for the detection of fecal residue but requires expertise to interpret. In order to be used by meat cutters without special training, automated detection is needed. This study used fluorescence imaging and deep learning algorithms to automatically detect and segment areas of fecal matter in carcass images using EfficientNet-B0 to determine which meat surface images showed fecal contamination and then U-Net to precisely segment the areas of contamination. The EfficientNet-B0 model achieved a 97.32% accuracy (precision 97.66%, recall 97.06%, specificity 97.59%, F-score 97.35%) for discriminating clean and contaminated areas on carcasses. U-Net segmented areas with fecal residue with an intersection over union (IoU) score of 89.34% (precision 92.95%, recall 95.84%, specificity 99.79%, F-score 94.37%, and AUC 99.54%). These results demonstrate that the combination of deep learning and fluorescence imaging techniques can improve food safety assurance by allowing the industry to use CSI-D fluorescence imaging to train employees in trimming carcasses as part of their Hazard Analysis Critical Control Point zero-tolerance plan. [ABSTRACT FROM AUTHOR]

Detection and classification of potato defects are of great significance to ensure food safety and improve product quality. This study investigated the potential of a novel multispectral imaging system based on single shot method for detection and classification of potato defects. A total of 417 potato samples were used in the experiment. The 25 spectral images with spatial resolution of 409 × 216 pixels over the spectral region between 676 and 952 nm were acquired for each potato. After improving the image contrast between the defect and defect-free regions by band math method, the defect regions were segmented from whole samples by using simple threshold. The spectral and textural features of the segmented regions were calculated and used for classification. A model for classifying different defects of potato was developed using least squares-support vector machine (LS-SVM) based on all feature set. The LS-SVM model achieved the classification accuracy of 90.70% for the test set. This research demonstrated that multispectral imaging system based on single shot method is a potential tool for online detection and classification of potato defects. The proposed data processing algorithm can be used for non-destructive testing of potato. [ABSTRACT FROM AUTHOR]

Spectroscopic techniques for food quality analysis are limited to surface inspections and are highly affected by the superficial layers (skin or packaging material) of the food samples. The ability of spatially offset Raman spectroscopy (SORS) to obtain chemical information from below the surface of a sample makes it a promising candidate for the non-destructive analysis of the quality of packaged food. In the present study, we developed a line-scan SORS technique for obtaining the Raman spectra of packaged-food samples. This technique was used to quantify butter adulteration with margarine through two different types of packaging. Further, the significant commercial potential of the developed technique was demonstrated by its being able to discriminate between ten commercial varieties of butter and margarine whilst still in their original, unopened packaging. The results revealed that, while conventional backscattering Raman spectroscopy cannot penetrate the packaging, thus preventing its application to the quality analysis of packaged food, SORS analysis yielded excellent qualitative and quantitative analyses of butter samples. The partial least-square regression analysis predictive values for the SORS data exhibit correlation coefficient values of 0.95 and 0.92, associated with the prediction error 3.2 % and 3.9 % for cover-1 & 2, respectively. The developed system utilizes a laser line (ca. 14-cm wide) that enables the simultaneous collection of a large number of spectra from a sample. Thus, by averaging the spectra collected for a given sample, the signal-to-noise ratio of the final spectrum can be enhanced, which will then have a significant effect on the multivariate data analysis methods used for qualitative and/or qualitative analyses. This recently presented line-scan SORS technique could be applied to the development of high-throughput and real-time analysis techniques for determining the quality and authenticity various packaged agricultural products. [ABSTRACT FROM AUTHOR]

Adulteration and fraud for powdered foods and ingredients are rising food safety risks that threaten consumers' health. In this study, a newly developed line-scan macro-scale Raman imaging system using a 5-W 785-nm line laser as excitation source was used to authenticate food powders. The system was used to collect hyperspectral Raman images in a wave number range of 102-2865 cm from three representative food powders mixed with selected chemical adulterants with a concentration of 0.5 %, including milk powder and melamine, wheat flour and benzoyl peroxide, and corn starch and maleic anhydride. An acoustic mixer that generates high-intensity acoustic waves was used to create food-adulterant mixtures. Maximum penetration depths of the line laser through powders of milk, flour, and starch were all estimated at 3 mm using layered food-on-adulterant samples. All the mixed samples were placed in sample holders with a surface area of 50 mm × 50 mm and a depth of 3 mm to ensure the adulterant particles at the very bottom can still be detected. Raman spectral and image processing algorithms were developed based on fluorescence correction and single-band images at unique Raman peaks of the individual adulterants. Chemical images were created to show identification, spatial distribution, and morphological features of the adulterant particles mixed in the food powders. The potential of estimating mass concentrations of the adulterants using the percentages of the adulterant pixels in the chemical images (i.e., volume concentrations) was also demonstrated. [ABSTRACT FROM AUTHOR]

Objective: The purpose of this study was to evaluate the impact of the negative lymph node (LN) count on the survival of the breast cancer patients in early stage after the axillary dissection. Methods: The breast cancer patients with TNM stage between January 2001 and December 2005 in Jiangsu Cancer Hospital, who underwent the axillary LNs dissection, were enrolled in this study. We analyzed the data of these patients including information of follow-up and postoperative pathological results. All patients were divided into two groups according to the axillary LN status and each group was divided into four subgroups according to the negative LN count. Cox regression analysis was performed to screen the pathological factor including the negative LN count on the survival and to compare the different negative LN count on the survival. Results: COX proportional hazard regression model showed that the survival of the breast cancer was significantly associated with the negative LN count. In TN group, when the negative LN count was 3 or less, 4 to 5, 6 to 9 and 10 or more, the median survival time was (82.6 ± 4.1) months, (101.5 ± 1.3) months, (104.7 ± 1.0) months, and (110.5 ± 0.9) months respectively ( P < 0.05). In TN group, when the negative LN count was 6 or less, 7 to 8, 9 to 10 and 11 or more, the median survival time was (95.4 ± 1.9) months, (101.8 ± 1.1) months, (104.9 ± 1.0) months, and (106.5 ± 0.9) months respectively ( P< 0.05). Conclusion: The negative LN count can reflect the adequacy of the axillary dissection. Increasing negative LN count is independently associated with improved survival in pTNM or pTNM staging breast cancer patients. The negative LN count should be considered for incorporation into staging for breast cancer with the axillary LN dissection. [ABSTRACT FROM AUTHOR]

The citrus industry has need for effective and efficient approaches to remove fruit with canker before they are shipped to selective international markets. The objective of this research was to study the effect of fruit harvest time on citrus canker detection using hyperspectral reflectance imaging. Ruby Red grapefruits with normal surface, canker, and five common peel diseases including greasy spot, insect damage, melanose, scab, and wind scar were collected during a 7-month harvest period. Hyperspectral reflectance images were acquired in the wavelength range of 450-930 nm. Spectral information divergence (SID) was used as a discrimination measure to perform statistical comparisons among reflectance spectra of grapefruit samples over the whole harvest season. The SID values with respect to the mean reflectance spectrum of canker for the 7-month periods were 0.0009, 0.0002, 0.0008, 0.0001, 0.0007, 0.0003, and 0.0004, respectively. Correlation analysis (CA) was used for hyperspectral band selection. Two-band ratio images using wavelengths of 729 and 834 nm selected by CA (R834/R729) gave the maximum absolute correlation value of 0.811. The mean ratio values for canker were in the range from 1.287 to 1.407, which were higher than the ratio values for other peel conditions. A simple thresholding and morphological filtering operations were applied to the two-band ratio images. The classification accuracies were in the range of 93.3-96.7% for each month. The results presented in this study demonstrated that there is no significant difference among the accuracy for canker detection over the whole harvest season using the two-band ratio images and threshold based on the spectrum of 7-month average. [ABSTRACT FROM AUTHOR]

RecQ helicases play a central role in maintaining genome stability and may interact with some important cancer-related proteins such as BRCA1. Mutations of the human RecQ helicase genes WRN and BLM lead to rare autosomal recessive disorders, Werner and Bloom syndromes, which are associated with premature aging and cancer predisposition, including breast cancer. In this case–control study of 1,004 breast cancer cases and 1,008 controls, we tested the hypothesis that non-conservative amino acid exchanges in WRN (leu1074Phe), BLM (Met298Thr) and BRCA1 (Pro871Leu) are independently or jointly associated with the risk of breast cancer in Chinese women. We found that the variant genotype of WRN Leu1074Phe was associated with a 1.36-fold significantly increased risk of breast cancer (OR = 1.36, 95% CI = 1.06–1.74). Moreover, a significant gene–environment interaction was evident between WRN leu1074Phe and age at menarche ( Pint = 0.02). Subjects carrying Phe/Phe genotype and with earlier age at menarche had 3.58-fold increased risk of breast cancer (OR = 3.58, 95% CI = 2.54–5.05). However, we did not find the significant main effect of polymorphisms in BLM and BRCA1 and also no locus–locus interactions were identified between WRN, BLM and BRCA1. These findings indicate that WRN leu1074Phe variant may contribute to the susceptibility of breast cancer in Chinese women. [ABSTRACT FROM AUTHOR]

Citrus canker is one of the most devastating diseases that threaten marketability of citrus crops. Technologies that can efficiently identify citrus canker would assure fruit quality and safety and enhance the competitiveness and profitability of the citrus industry. This research was aimed to investigate the potential of using hyperspectral imaging technique for detecting canker lesions on citrus fruit. A portable hyperspectral imaging system consisting of an automatic sample handling unit, a light source, and a hyperspectral imaging unit was developed for citrus canker detection. The imaging system was used to acquire reflectance images from citrus samples in the wavelength range between 400 and 900 nm. Ruby Red grapefruits with normal and various diseased skin conditions including canker, copper burn, greasy spot, wind scar, cake melanose, and specular melanose were tested. Hyperspectral reflectance images were analyzed using principal component analysis (PCA) to compress the 3-D hyperspectral image data and extract useful image features that could be used to discriminate cankerous samples from normal and other diseased samples. Image processing and classification algorithms were developed based upon the transformed images of PCA. The overall accuracy for canker detection was 92.7%. Four optimal wavelengths (553, 677, 718, and 858 nm) were identified in visible and short-wavelength near-infrared region that could be adopted by a future multispectral imaging solution for detecting citrus canker on a sorting machine. This research demonstrated that hyperspectral imaging technique could be used for discriminating citrus canker from other confounding diseases. [ABSTRACT FROM AUTHOR]