RESUMEN
BACKGROUND: Automatic skin lesion recognition has shown to be effective in increasing access to reliable dermatology evaluation; however, most existing algorithms rely solely on images. Many diagnostic rules, including the 3-point checklist, are not considered by artificial intelligence algorithms, which comprise human knowledge and reflect the diagnosis process of human experts. OBJECTIVE: In this paper, we aimed to develop a semisupervised model that can not only integrate the dermoscopic features and scoring rule from the 3-point checklist but also automate the feature-annotation process. METHODS: We first trained the semisupervised model on a small, annotated data set with disease and dermoscopic feature labels and tried to improve the classification accuracy by integrating the 3-point checklist using ranking loss function. We then used a large, unlabeled data set with only disease label to learn from the trained algorithm to automatically classify skin lesions and features. RESULTS: After adding the 3-point checklist to our model, its performance for melanoma classification improved from a mean of 0.8867 (SD 0.0191) to 0.8943 (SD 0.0115) under 5-fold cross-validation. The trained semisupervised model can automatically detect 3 dermoscopic features from the 3-point checklist, with best performances of 0.80 (area under the curve [AUC] 0.8380), 0.89 (AUC 0.9036), and 0.76 (AUC 0.8444), in some cases outperforming human annotators. CONCLUSIONS: Our proposed semisupervised learning framework can help with the automatic diagnosis of skin disease based on its ability to detect dermoscopic features and automate the label-annotation process. The framework can also help combine semantic knowledge with a computer algorithm to arrive at a more accurate and more interpretable diagnostic result, which can be applied to broader use cases.
RESUMEN
The surface-enhanced activities of size- and shape-controlled gold nanoparticles (AuNPs) with superior chemical stability were investigated to explore a possible development of a simple and non-destructive spectroscopic method to help the regulatory agency's analytical services for rapid detection and characterization of selected antimicrobials in animal feeds. Feed samples spiked at different concentration ranges of antimicrobials were evaluated using AuNPs as a surface-enhanced Raman spectroscopy (SERS) agent. The collected SERS spectra were mathematically preprocessed for further analysis. The classification models obtained 100% predictive accuracy with zero or little misclassification. The first two canonical variables (p = 0.001) could explain >95% of the variability in preprocessed spectral data. Most chemometric models for predicting MON, DEC, and LAS concentrations showed a high predictive accuracy (r2 > 0.90), lower predictive error (<20 mg/kg), and satisfactory regression quality (slope close to 1.0). The statistical results showed no statistically significant difference between the reference and SERS predicted values (p > 0.05). The findings and implications from the study indicate that SERS would be a powerful and efficient technique possessing a great potential serving as an excellent monitoring and screening tool for antimicrobial contaminated samples in the on-site analysis.