This study developed an automated temperature measurement and monitoring platform for dairy cattle. The platform used the YOLO V3-tiny (you only look once, YOLO) deep learning algorithm to identify and classify dairy cattle images. The system included a total of three layers of YOLO V3-tiny identification: (1) dairy cow body; (2) individual number (identity, ID); (3) thermal image of eye socket identification. We recorded each cow’s individual number and body temperature data after the three layers of identification, and carried out long-term body temperature tracking. The average prediction score of the recognition rate was 96%, and the accuracy was 90.0%. The thermal image of eye socket recognition rate was >99%. The area under the receiver operating characteristic curves (AUC) index of the prediction model was 0.813 (0.717–0.910). This showed that the model had excellent predictive ability. This system provides a rapid and convenient temperature measurement solution for ranchers. The improvement in dairy cattle image recognition can be optimized by collecting more image data. In the future, this platform is expected to replace the traditional solution of intrusive radio-frequency identification for individual recognition.

Vibroarthrographic (VAG) signals are used as useful indicators of knee osteoarthritis (OA) status. The objective was to build a template database of knee crepitus sounds. Internships can practice in the template database to shorten the time of training for diagnosis of OA. A knee sound signal was obtained using an innovative stethoscope device with a goniometer. Each knee sound signal was recorded with a Kellgren–Lawrence (KL) grade. The sound signal was segmented according to the goniometer data. The signal was Fourier transformed on the correlated frequency segment. An inverse Fourier transform was performed to obtain the time-domain signal. Haar wavelet transform was then done. The median and mean of the wavelet coefficients were chosen to inverse transform the synthesized signal in each KL category. The quality of the synthesized signal was assessed by a clinician. The sample signals were evaluated using different algorithms (median and mean). The accuracy rate of the median coefficient algorithm (93 %) was better than the mean coefficient algorithm (88 %) for cross-validation by a clinician using synthesis of VAG. The artificial signal we synthesized has the potential to build a learning system for medical students, internships and para-medical personnel for the diagnosis of OA. Therefore, our method provides a feasible way to evaluate crepitus sounds that may assist in the diagnosis of knee OA.

The neodymium-doped yttrium aluminum garnet (Nd-YAG) laser is used for removal of pigmented skin patches and rejuvenation of skin. However, complications such as hyperpigmentation, hypopigmentation, and petechiae can occur after frequent treatments. Therefore, identifying the risk factors for such complications is important. The development of a multivariable logistic regression model with least absolute shrinkage and selection operator (LASSO) is needed to provide valid predictions about the incidence of post inflammatory hyperpigmentation complication probability (PIHCP) among patients treated with Nd-YAG laser toning. A total of 125 female patients undergoing laser toning therapy between January 2014 and January 2016 were examined for post-inflammatory hyperpigmentation (PIH) complications. Factor analysis was performed using 15 potential predictive risk factors of PIH determined by a physician. The LASSO algorithm with cross-validation was used to select the optimal number of predictive risk factors from the potential factors for a multivariate logistic regression PIH complication model. The optimal number of predictive risk factors for the model was five: immediate endpoints of laser (IEL), α-hydroxy acid (AHA) peels, Fitzpatrick skin phototype (FSPT), acne, and melasma. The area under the receiver operating characteristic curve (AUC) was 0.79 (95% CI, 0.70–0.88) in the optimal model. The overall performance of the LASSO-based PIHCP model was satisfactory based on the AUC, Omnibus, Nagelkerke R², and Hosmer–Lemeshow tests. This predictive risk factor model is useful to further optimize laser toning treatment related to PIH. The LASSO-based PIHCP model could be useful for decision-making.

Radiation-induced pneumonitis (RP) is a common complication in breast cancer patients after radiation therapy (RT). In the present study, least absolute shrinkage and selection operator (LASSO) and five classification algorithms were used to improve both the predictive ability of RP and the quality of patients’ daily life. A total of 106 breast cancer patients were enrolled in this study. All of the patients were treated with volumetric modulated arc therapy (VMAT). A total of 19 risk factors were included in this study. The present study found that the area under receiver operating characteristics curve (AUC) and accuracy (ACC) of LASSO selected factors (F_LASSO) for each of the five classification algorithms were generally higher than those of all selected factors (F_all) and dose selected factors (F_dose). We propose to use LASSO with support vector machine (SVM) to assess the risk of complications, to improve the predictive ability for breast cancer patients with complications after RT, and to reduce the cost of assessing the risk of complications.

Illumination variation is among the several bottlenecks in a face recognition system because it can greatly affect the appearance of a face image, which causes a reduction in the face matching performance of the system. Using illumination preprocessing methods is an effective approach to overcome this problem. Despite the achievements made, however, each method still has its own demerits. In this paper, an efficient representation method insensitive to varying illumination based on a combination of Weber-face (WF) and the difference of Gaussians (DoG) methods is proposed for human face recognition using local descriptors. After processing by our method, the obtained image will preserve more facial features and edge information while shading effects are eliminated. To demonstrate the potential of the proposed method, two systems using nearest-neighbor and support vector machine classifiers were implemented. Experimental results for the CMU-PIE and extended Yale B face databases showed that our method could effectively eliminate the effect of uneven illumination and achieved better recognition accuracies in comparison to other state-of-the-art methods (DoG, gradient face, histogram equalization, and WF).

Computed Tomography (CT) technology has been widely used for detecting and diagnosing lung disease. To improve the visibility of essential features in a CT image, reducing noise and blur, sharpening features, and increasing contrast are necessary. In this study, we propose a method to improve these features and gain better characteristics of region of interest in lung CT images for a right diagnosis. Processing of the proposed method consists of non-local means filter for removing noise, unsharp masking for deblurring and sharpening image, and finally singular value decomposition for enhancing contrast of the region of interest. The method was evaluated in terms of improvement in contrast based on contrast improvement ratio in lung CT images. The results clearly demonstrated that our method reached a higher contrast improvement ratio compared to histogram equalization, unsharp masking, contrast-limited adaptive histogram equalization, and singular value equalization methods and helped to increase the clarity of relevant details without distorting the images.