Semi-supervised learning for industrial fault detection and diagnosis: A systemic review.

The automation of Fault Detection and Diagnosis (FDD) is a central task for many industries today. A myriad of methods are in use, although the most recent leading contenders are data-driven approaches and especially Machine Learning (ML) methods. ML algorithms fall into two main categories: supervised and unsupervised methods, depending on whether or not the instances are labeled with the expected outputs. However, a new approach called Semi-Supervised Learning (SSL) has recently emerged that uses a few labeled instances together with other unlabeled instances for the training process. This new approach can significantly improve the accuracy of conventional ML models for industrial environments where labeled data are scarce. SSL has been tested as a promising solution over the past few years for several FDD problems, although there have been no systemic reviews of this sort of approach up until the present review. In this study, an attempt to organize the existing literature on SSL for FDD using the taxonomy of van Engelen & Hoos is reported. The most and the least frequently used SSL algorithms are identified and considered in terms of different fault detection tasks and their most common dataset structure. Moreover, a set of best practices are proposed in the conclusions of this work for implementation under real industrial conditions, so as to avoid some of the most common faults.

Virtual Reality for Learning.

The consumer age of the Personal Computer and mobile devices has opened up a new world of opportunities for innovative teaching methodologies, many based on serious games and virtual worlds. Similar levels of market penetration are expected for the use of Immersive Virtual Reality (iVR) over upcoming decades, once all the core technologies for game engines and head-mounted displays are available on the market at affordable prices. In this chapter, a general overview of the state of the art of iVR learning experiences is presented. Firstly, the advantages of iVR over traditional learning are described - advantages that must be considered when defining iVR experiences for the optimization of student learning and satisfaction. Secondly, the relationship between learning theories and iVR experiences is briefly summarized; an area where constructivist theories appear to be the most commonly used theory in iVR experiences. Thirdly, some examples of the success of iVR applications at different learning levels, from primary school to higher education, are summarized. Fourthly, the key factors for the successful design and use of an iVR experience in education are identified, from the predesign stage to the final evaluation - with special attention given to the different possibilities of each type of HMD for different kinds of educational experiences. Finally, the main limitations of iVR for learning today and the future trends of this technology for teaching are also identified and discussed.

Immersive virtual-reality computer-assembly serious game to enhance autonomous learning.

Immersive virtual reality (VR) environments create a very strong sense of presence and immersion. Nowadays, especially when student isolation and online autonomous learning is required, such sensations can provide higher satisfaction and learning rates than conventional teaching. However, up until the present, learning outcomes with VR tools have yet to prove their advantageous aspects over conventional teaching. The project presents a VR serious game for teaching concepts associated with computer hardware assembly. These concepts are often included in any undergraduate's introduction to Computer Science. The learning outcomes are evaluated using a pre-test of previous knowledge, a satisfaction/usability test, and a post-test on knowledge acquisition, structured with questions on different knowledge areas. The results of the VR serious game are compared with another two learning methodologies adapted to online learning: (1) an online conventional lecture; and (2) playing the same serious game on a desktop PC. An extensive sample of students (n = 77) was formed for this purpose. The results showed the strong potential of VR serious games to improve student well-being during spells of confinement, due to higher learning satisfaction. Besides, ease of usability and the use of in-game tutorials are directly related with game-user satisfaction and performance. The main novelty of this research is related to academic performance. Although a very limited effect was noted for learning theoretical knowledge with the VR application in comparison with the other methodologies, this effect was significantly improved through visual knowledge, understanding and making connections between different concepts. It can therefore be concluded that the proposed VR serious game has the potential to increase student learning and therefore student satisfaction, by imparting a deeper understanding of the subject matter to students. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s10055-021-00607-1.

Awareness, Prevention, Detection, and Therapy Applications for Depression and Anxiety in Serious Games for Children and Adolescents: Systematic Review.

BACKGROUND: Depression and anxiety in children and adolescents are major health problems worldwide. In recent years, serious games research has advanced in the development of tools to address these mental health conditions. However, there has not been an extensive analysis of these games, their tendencies, and capacities. OBJECTIVE: This review aims to gather the most current serious games, published from 2015 to 2020, with a new approach focusing on their applications: awareness, prevention, detection, and therapy. The purpose is also to analyze the implementation, development, and evaluation of these tools to obtain trends, strengths, and weaknesses for future research lines. METHODS: The identification of the serious games through a literature search was conducted on the databases PubMed, Scopus, Wiley, Taylor and Francis, Springer, PsycINFO, PsycArticles, Web of Science, and Science Direct. The identified records were screened to include only the manuscripts meeting these criteria: a serious game for PC, smartphone, or virtual reality; developed by research teams; targeting only depression or anxiety or both; aiming specifically at children or adolescents. RESULTS: A total of 34 studies have been found that developed serious games for PC, smartphone, and virtual reality devices and tested them in children and adolescents. Most of the games address both conditions and are applied in prevention and therapy. Nevertheless, there is a trend that anxiety is targeted more in childhood and depression targeted more in adolescence. Regarding design, the game genres arcade minigames, adventure worlds, and social simulations are used, in this order. For implementation, these serious games usually require sessions of 1 hour and are most often played using a PC. Moreover, the common evaluation tools are normalized questionnaires that measure acquisition of skills or reduction of symptoms. Most studies collect and compare these data before and after the participants play. CONCLUSIONS: The results show that more awareness and detection games are needed, as well as games that mix the awareness, prevention, detection, and therapy applications. In addition, games for depression and anxiety should equally target all age ranges. For future research, the development and evaluation of serious games should be standardized, so the implementation of serious games as tools would advance. The games should always offer support while playing, in addition to collecting data on participant behavior during the game to better analyze their learning. Furthermore, there is an open line regarding the use of virtual reality for these games due to the capabilities offered by this technology.

A regression-tree multilayer-perceptron hybrid strategy for the prediction of ore crushing-plate lifetimes.

Highly tensile manganese steel is in great demand owing to its high tensile strength under shock loads. All workpieces are produced through casting, because it is highly difficult to machine. The probabilistic aspects of its casting, its variable composition, and the different casting techniques must all be considered for the optimisation of its mechanical properties. A hybrid strategy is therefore proposed which combines decision trees and artificial neural networks (ANNs) for accurate and reliable prediction models for ore crushing plate lifetimes. The strategic blend of these two high-accuracy prediction models is used to generate simple decision trees which can reveal the main dataset features, thereby facilitating decision-making. Following a complexity analysis of a dataset with 450 different plates, the best model consisted of 9 different multilayer perceptrons, the inputs of which were only the Fe and Mn plate compositions. The model recorded a low root mean square error (RMSE) of only 0.0614 h for the lifetime of the plate: a very accurate result considering their varied lifetimes of between 746 and 6902 h in the dataset. Finally, the use of these models under real industrial conditions is presented in a heat map, namely a 2D representation of the main manufacturing process inputs with a colour scale which shows the predicted output, i.e. the expected lifetime of the manufactured plates. Thus, the hybrid strategy extracts core training dataset information in high-accuracy prediction models. This novel strategy merges the different capabilities of two families of machine-learning algorithms. It provides a high-accuracy industrial tool for the prediction of the full lifetime of highly tensile manganese steel plates. The results yielded a precision prediction of (RMSE of 0.061 h) for the full lifetime of (light, medium, and heavy) crusher plates manufactured with the three (experimental, classic, and highly efficient (new)) casting methods.

An SVM-based solution for fault detection in wind turbines.

Research into fault diagnosis in machines with a wide range of variable loads and speeds, such as wind turbines, is of great industrial interest. Analysis of the power signals emitted by wind turbines for the diagnosis of mechanical faults in their mechanical transmission chain is insufficient. A successful diagnosis requires the inclusion of accelerometers to evaluate vibrations. This work presents a multi-sensory system for fault diagnosis in wind turbines, combined with a data-mining solution for the classification of the operational state of the turbine. The selected sensors are accelerometers, in which vibration signals are processed using angular resampling techniques and electrical, torque and speed measurements. Support vector machines (SVMs) are selected for the classification task, including two traditional and two promising new kernels. This multi-sensory system has been validated on a test-bed that simulates the real conditions of wind turbines with two fault typologies: misalignment and imbalance. Comparison of SVM performance with the results of artificial neural networks (ANNs) shows that linear kernel SVM outperforms other kernels and ANNs in terms of accuracy, training and tuning times. The suitability and superior performance of linear SVM is also experimentally analyzed, to conclude that this data acquisition technique generates linearly separable datasets.

A virtual sensor for online fault detection of multitooth-tools.

The installation of suitable sensors close to the tool tip on milling centres is not possible in industrial environments. It is therefore necessary to design virtual sensors for these machines to perform online fault detection in many industrial tasks. This paper presents a virtual sensor for online fault detection of multitooth tools based on a bayesian classifier. The device that performs this task applies mathematical models that function in conjunction with physical sensors. Only two experimental variables are collected from the milling centre that performs the machining operations: the electrical power consumption of the feed drive and the time required for machining each workpiece. The task of achieving reliable signals from a milling process is especially complex when multitooth tools are used, because each kind of cutting insert in the milling centre only works on each workpiece during a certain time window. Great effort has gone into designing a robust virtual sensor that can avoid re-calibration due to, e.g., maintenance operations. The virtual sensor developed as a result of this research is successfully validated under real conditions on a milling centre used for the mass production of automobile engine crankshafts. Recognition accuracy, calculated with a k-fold cross validation, had on average 0.957 of true positives and 0.986 of true negatives. Moreover, measured accuracy was 98%, which suggests that the virtual sensor correctly identifies new cases.

Deep plane face-lift.

The deep plane face-lift is a safe and reliable technique for treatment of the aging face. It affords improvement in the lower face and midface. The use of fibrin glue in the face-lift has improved the healing and recovery of patients.

The extended columellar strut-tip graft.

The extended columellar strut-tip graft is a structural unit used in endonasal rhinoplasty that combines the attributes of the columellar strut and the tip graft. It is used to provide projection and contour to the nasal tip. Our goal with this study was to evaluate a 15-year experience with 155 patients who underwent rhinoplasty with the extended columellar strut-tip graft. Of these, 110 underwent secondary rhinoplasty, and 45 underwent primary rhinoplasty. There were 6 patients in the secondary rhinoplasty group who experienced complications: in 3, the graft became visible postoperatively, and 3 patients had graft placement asymmetry. These 6 patients underwent surgery in the initial years of graft development. One patient with graft edge visibility and 1 patient with graft asymmetry underwent revision surgery with satisfactory results. The extended columella strut-tip graft is a reliable method to provide nasal tip projection and contour. The successful use of the graft requires precise diagnosis and surgical technique.

Nasal valve surgery improves disease-specific quality of life.

OBJECTIVES/HYPOTHESIS: Disease-specific quality-of-life (QOL) assessment of patients with nasal valve compromise and symptomatic nasal obstruction has not been studied previously. The objectives of the study were to determine whether surgical treatment of the nasal valve improves disease-specific QOL and to identify clinical or demographic variables predictive of patients' baseline QOL or change in QOL. STUDY DESIGN: Prospective, multi-institutional outcomes study of 20 patients with nasal obstruction and a surgically treatable diagnosis of nasal valve compromise. METHODS: Disease-specific QOL assessment was performed using the Nasal Obstruction Symptom Evaluation (NOSE) scale preoperatively (n = 20) and at 3 (n = 14) and 6 months (n = 20) after surgery. Clinical and demographic data were collected, along with physician-reported assessments of degree of nasal obstruction. RESULTS: Mean NOSE scores significantly improved from baseline to 3 months after surgery (68.9 vs. 20.7 [P < .0001]), from baseline to 6 months after surgery (68.9 vs. 15.8 [P < .0001]), and from 3 to 6 months after surgery (20.7 vs. 15.8 [P = .0077]). Physician assessment of degree of nasal obstruction using a visual analogue scale was significantly correlated with baseline NOSE scores (P = .013) and change in NOSE scores at 6 months (P = .0015). No other clinical or demographic factors were found to be predictive. CONCLUSION: In patients with symptomatic nasal obstruction and nasal valve compromise, surgical repair of the nasal valve improves disease-specific QOL. Physician rating of degree of nasal obstruction was found to be significantly correlated with patient-reported QOL.

Revision single-stage laryngotracheal reconstruction in children.

In this report, we discuss indications, technique, outcome, and complications of revision single-stage laryngotracheal reconstruction (SSLTR), formulate guidelines to avoid or prevent procedure failure, and establish a protocol for the management of procedure failure. We retrospectively reviewed the charts of 122 patients between the ages of 8 months and 9 years who underwent SSLTR between January 1992 and September 2001 in 2 tertiary care children's medical centers in different cities and assessed the outcomes of patients who underwent revision SSLTR. A total of 122 patients underwent SSLTR, of whom 48 patients underwent anterior and posterior grafting. Of the 122 patients, 13 had revision SSLTR; 8 of these 13 underwent the initial laryngotracheal reconstruction at another institution. Five patients had anterior grafting laryngotracheal reconstruction without stenting, 7 had anterior and posterior grafting with 1 to 21 days of endotracheal intubation, and I had cricotracheal resection and anastomosis. Of the 13 patients, 5 had anterior wall or graft collapse (grade IV stenosis), 4 had subglottic stenosis (grade IV), 2 had circumferential subglottic stenosis (grade III), and 2 had subglottic and glottic stenosis (grade IV). The overall success rate for all patients was 86% (105 of 122). The success rates for the 122 patients were as follows: anterior grafting, 100%; anterior and posterior grafting, 83% (40 of 48); and revision cases, 70% (9 of 13). We conclude that laryngotracheal reconstruction with a costal cartilage rib graft should be considered the procedure of choice for the management of subglottic stenosis. We believe that patients in whom the first procedure fails should have a high chance of success with revision SSLTR if strict guidelines and protocols are followed.

Octreotide scintigraphy in the detection of recurrent paragangliomas.

OBJECTIVE: Conventional follow-up of surgically and radiotherapy-treated paragangliomas is usually carried out with physical examinations and either magnetic resonance imaging and/or computed tomography. Up to now, octreotide has been reserved for primary diagnosis, for which it has been shown to be a reliable imaging modality. However, the postsurgical and postradiation changes seen on conventional imaging make the diagnosis of small recurrent lesions difficult. We sought to evaluate the role of octreotide scintigraphy in the detection of recurrent paragangliomas. STUDY DESIGN: We based this case series on 3 patients who were found to have recurrent paragangliomas after primary surgical resection. All patients underwent octreotide scintigraphy during routine follow-up. RESULTS: Three patients were found to have recurrent paragangliomas using octreotide scintigraphy. In one of the patients, conventional imaging (magnetic resonance imaging, computed tomography) failed to detect the recurrence. CONCLUSION: The octreotide study was able to detect the small recurrences because octreotide binds directly to receptors on the tumor and is not affected by postsurgical or postradiotherapy changes. The intrinsic resolution of most commercially available gamma cameras is about 1 cm, making it quite sensitive to detect small tumors. Although this patient did not have synchronous lesions, octreotide scintigraphy can also aide in the detection of such lesions.

Efficacy of perioperative thromboprophylactic agents in the maintenance of anastamotic patency and survival of rat microvascular free groin flaps.

OBJECTIVES: We sought to determine the thromboprophylactic effects of low-molecular-weight heparin (LMWH) and pentoxifylline. DESIGN: Randomized blinded study. Materials and methods Sixty Sprague-Dawley rats were assigned to 1 of 4 treatment groups: (1) control group (n = 15), (2) pentoxifylline group (n = 15), (3) LMWH group (n = 15), and (4) combined therapy group (n = 15). A groin/abdominal free flap was elevated in each, with division of the artery and microvascular repair The animals received therapy for 4 days preoperatively and 6 days postoperatively. Flaps were monitored visually, and on the seventh postoperative day, the flap was elevated, and the microvascular anastomosis was assessed for patency. RESULTS: There was a statistically significant improvement in arterial patency in both pentoxifylline and LMWH groups compared with controls (P < 0.01) but no significant improvement with combined therapy (P > 0.05). CONCLUSIONS: In this animal model, LMWH and pentoxifylline have significant thrombroprophylactic effects when used individually but not in combination.

Orbital infection as a complication of sinusitis: are diagnostic and treatment trends changing?

Orbital infection has long been the most common complication of sinusitis. In light of our increased knowledge of sinusitis, improved diagnostic tools, and new pharmacologic and surgical treatments, we investigated whether trends in diagnosis and treatment are changing. We reviewed the charts of all 43 patients who had been referred to our institution with orbital complications of sinusitis between Jan. 1, 1985, and Dec. 31, 1999. Nine of the 43 patients had been diagnosed between Jan. 1, 1985, and Dec. 31, 1990 (mean: 1.5 patients/yr) and 34 had been diagnosed between Jan. 1, 1991, and Dec. 31, 1999 (mean: 3.8 patients/yr). Of the 43 patients, 27 had cellulitis and 16 had an abscess (one of the 16 had two abscesses--one subperiosteal and one supraorbital). All 17 abscesses were treated surgically. Five of the 7 abscesses operated on from 1985 through 1990 were treated via an open external approach, whereas 7 of the 10 abscesses that were operated on later were treated via an endoscopic approach. We conclude that orbital complications of sinonasal origin are being recognized more frequently than they were in the past and that endoscopy has supplanted the open external approach as the preferred method of drainage.