Excessive radiographic exposure in the follow-up of adolescent idiopathic scoliosis (AIS) remains a clinical concern. Surface topography (ST) and angle of trunk rotation (ATR) have shown promise for non-radiographic monitoring, although their ability to detect clinically meaningful Cobb angle changes (> 5°) remains limited. This study aimed to validate a machine learning model for predicting Cobb angle progression using ST parameters and ATR obtained at three and six months. A prospective observational study was conducted in 43 AIS patients (57 curves) recruited from two centers. Baseline and six-month radiographic Cobb angles were recorded along with ATR and five ST asymmetry parameters (MaxDev, RMS, LatDev, hump volume, asymmetry patch area). A random forest (RF) model was used to predict Cobb angles at 3 and 6 months and then to estimate progression over 6 months (ΔCobb). Outcomes were classified as improvement (ΔCobb < -5°), stabilization (-5° ≤ ΔCobb ≤ + 5°), or progression (ΔCobb > + 5°). The RF model predicted the six-month radiographic Cobb with MAE of 7.03° (three-month input) and 6.91° (six-month input). The progression model integrating both time points achieved an overall accuracy of 80.7
Die Sicherheit in Maschinenräumen auf Kreuzfahrtschiffen ist aufgrund komplexer Layouts, hoher Personenzahlen und weitläufiger Versorgungssysteme eine Herausforderung. Notfälle wie Dampfrohrbrüche können die Sicht stark einschränken, wodurch es zu Verzögerungen bei der Evakuierung in gefährdeten Bereichen kommt. Dies verzögert die Erkennung hilfsbedürftiger Personen und erhöht das Verletzungsrisiko. Wir präsentieren ein KI-basiertes Proof of Concept zur Echtzeit-Personenerkennung und -lokalisierung. Bei diesem Ansatz werden Überwachungskameras genutzt welche Personen erfasst und die ermittelten Standorte auf einer Karte mit vordefinierten Schiffsbereichen visualisiert. Durch zeitliche Verfolgung wird eine aktuelle Übersicht über die Personenverteilung gewährleistet. Das System wurde in einer simulierten Indoor-Umgebung getestet. Raspberry Pi-Kameras, synchronisiert über das Network Time Protocol (NTP), stellen einheitliche Zeitstempel sicher. Überlappende Kamerabereiche wurden optimiert, um Mehrfachzählungen zu vermeiden. Zur Personenerkennung wurde das vortrainierte Echtzeitmodell YOLOv8 verwendet. Die Standorte wurden mittels OpenCV und Matplotlib auf einer Modell-Karte visualisiert. Zukünftige Arbeiten fokussieren auf verbesserte Robustheit bei schlechter Sicht, Skalierbarkeit für größere Überwachungsbereiche und optimierte Modellierung. Über Kreuzfahrtschiffe hinaus bietet das System Potenzial für andere sicherheitskritische Bereiche mit hoher Personendichte und komplexen Infrastrukturen. Die präzise Echtzeitverfolgung erleichtert und beschleunigt Rettungseinsätze und erhöht Sicherheitsstandards für Passagiere und Personal.
In failure analysis, micro-fractographic analysis of fracture surfaces is usually performed based on practical knowledge which is gained from available studies, own comparative tests, from the literature, as well as online databases. Based on comparisons with already existing images, fracture mechanisms are determined qualitatively. These images are mostly two-dimensional and obtained by light optical and scanning electron imaging techniques. So far, quantitative assessments have been limited to macroscopically determined percentages of fracture types or to the manual measurement of fatigue striations, for example. Recently, more and more approaches relying on computer algorithms have been taken, with algorithms capable of finding and classifying differently structured fracture characteristics. For the Industrial Collective Research (Industrielle Gemeinschaftsforschung, IGF) project "iFrakto " presented in this paper, electron-optical images are obtained, from which topographic information is calculated. This topographic information is analyzed together with the conventional 2D images. Analytical algorithms and deep learning are used to analyze and evaluate fracture characteristics and are linked to information from a fractography database. The most important aim is to provide software aiding in the application of fractography for failure analysis. This paper will present some first results of the project.
The finishing of yarns by bobbin dyeing is of key importance in the value chain of home and apparel textiles. In the bobbin dyeing process, yarn bobbins are placed on dye spindles and passed through with dye liquor. If the winding process parameters are selected unfavorably, the density of the winding package is inhomogeneous. The inhomogeneous density distribution leads to an inhomogeneous flow of dye liquid through the package and, hence, to dyeing defects. In order to minimize dyeing defects and to reduce set-up times, we present a simulation-based parameter recommendation for cross-winding machines in this paper. We use a kinematic model of the winding process combined with an empirical model for the package diameter to optimize the package density distribution. We introduce a criterion to avoid patterning defects and adapt winding settings. For bobbins with Nm 34 Co yarn, the homogeneity of the density was improved and the color deviation was reduced by up to 50% due to these simulation-based setting suggestions.
Patients with life-threatening heart failure where all conservative therapeutic options have been exhausted may be indicated for mechanical pulsatile ventricular assist devices. These paracorporeal VADs are used for short- and long-term support of left and/or right ventricular pumping function (see [1]). Currently, these systems require short monitoring cycles by clinical professionals and allow little mobility for the patient. To extend these cycles and improve mobility, this paper presents a method to detect increased risk of complications during the use of a VAD. The VADs membrane is monitored using acoustic measurements, where ultrasonic pulses are emitted and the corresponding echo is measured. The main challenge is that a change in the essential performance characteristics of the blood pump would threaten the already granted approval as a medical device. Furthermore, there is the problem that the ultrasonic pulse has to pass through a two-meter long tube and that the motion behavior of the membrane is not symmetrical, but rather resembles the inflation and deflation of a plastic bag. A supervised classification using support vector machine (SVM) has proven to be a sufficiently accurate method for this problem and this type of data. The SVM operates on the frequency spectra of the impulse responses and classifies three trained states that must occur during a single pump cycle if the system is fully functional. A faulty state of the system is detected by the absence of certain states within a pumping cycle.
