arXiv:2511.17219v1 Announce Type: new 
Abstract: The paper introduces DelTriC (Delaunay Triangulation Clustering), a clustering algorithm which integrates PCA/UMAP-based projection, Delaunay triangulation, and a novel back-projection mechanism to form clusters in the original high-dimensional space. DelTriC decouples neighborhood construction from decision-making by first triangulating in a low-dimensional proxy to index local adjacency, and then back-projecting to the original space to perform robust edge pruning, merging, and anomaly detection. DelTriC can outperform traditional methods such as k-means, DBSCAN, and HDBSCAN in many scenarios; it is both scalable and accurate, and it also significantly improves outlier detection.

يمثل تقديم DelTriC، وهو خوارزمية جديدة للتجميع، تقدمًا كبيرًا في تحليل البيانات من خلال دمج الإسقاط القائم على PCA/UMAP والتثليث ديلوني. تعزز هذه الطريقة تشكيل المجموعات في الفضاءات عالية الأبعاد بينما تحسن من اكتشاف القيم الشاذة وقابلية التوسع مقارنةً بالتقنيات التقليدية مثل k-means وDBSCAN.

La introducción de DelTriC, un nuevo algoritmo de agrupamiento, marca un avance significativo en el análisis de datos al integrar proyección basada en PCA/UMAP y triangulación de Delaunay. Este método mejora la formación de grupos en espacios de alta dimensión mientras mejora la detección de valores atípicos y la escalabilidad en comparación con técnicas tradicionales como k-means y DBSCAN.

L'introduction de DelTriC, un nouvel algorithme de clustering, marque une avancée significative dans l'analyse des données en intégrant la projection basée sur PCA/UMAP et la triangulation de Delaunay. Cette méthode améliore la formation de clusters dans des espaces de haute dimension tout en améliorant la détection des valeurs aberrantes et l'évolutivité par rapport aux techniques traditionnelles telles que k-means et DBSCAN.

The introduction of DelTriC, a novel clustering algorithm, marks a significant advancement in data analysis by integrating PCA/UMAP-based projection and Delaunay triangulation. This method enhances cluster formation in high-dimensional spaces while improving outlier detection and scalability compared to traditional techniques like k-means and DBSCAN.

DelTriC: A Novel Clustering Method with Accurate Outlier

arXiv:2511.17485v1 Announce Type: new 
Abstract: The human spine is a complex structure composed of 33 vertebrae. It holds the body and is important for leading a healthy life. The spine is vulnerable to age-related degenerations that can be identified through magnetic resonance imaging (MRI). In this paper we propose a novel computer-vison-based deep learning method to estimate spine age using images from over 18,000 MRI series. Data are restricted to subjects with only age-related spine degeneration. Eligibility criteria are created by identifying common age-based clusters of degenerative spine conditions using uniform manifold approximation and projection (UMAP) and hierarchical density-based spatial clustering of applications with noise (HDBSCAN). Model selection is determined using a detailed ablation study on data size, loss, and the effect of different spine regions. We evaluate the clinical utility of our model by calculating the difference between actual spine age and model-predicted age, the spine age gap (SAG), and examining the association between these differences and spine degenerative conditions and lifestyle factors. We find that SAG is associated with conditions including disc bulges, disc osteophytes, spinal stenosis, and fractures, as well as lifestyle factors like smoking and physically demanding work, and thus may be a useful biomarker for measuring overall spine health.

تم تطوير إطار جديد للذكاء الاصطناعي لقياس شيخوخة العمود الفقري البشري باستخدام التصوير بالرنين المغناطيسي، مستفيدًا من طريقة تعلم عميقة تحلل أكثر من 18,000 سلسلة من صور الرنين المغناطيسي التي تركز على تدهور العمود الفقري المرتبط بالعمر. يستخدم النموذج تقنيات تجميع متقدمة لتحديد الحالات التدهورية ويقيم فائدته السريرية من خلال مقارنة العمر الفعلي للعمود الفقري مع توقعات النموذج.

Se ha desarrollado un nuevo marco de inteligencia artificial para medir el envejecimiento de la columna vertebral humana utilizando resonancias magnéticas (IRM), aprovechando un método de aprendizaje profundo que analiza más de 18,000 series de IRM centradas en la degeneración espinal relacionada con la edad. El modelo emplea técnicas avanzadas de agrupamiento para identificar condiciones degenerativas y evalúa su utilidad clínica comparando la edad real de la columna con las predicciones del modelo.

Un nouveau cadre d'intelligence artificielle a été développé pour mesurer le vieillissement de la colonne vertébrale humaine à l'aide d'IRM, en s'appuyant sur une méthode d'apprentissage profond qui analyse plus de 18 000 séries d'IRM axées sur la dégénérescence vertébrale liée à l'âge. Le modèle utilise des techniques de clustering avancées pour identifier les conditions dégénératives et évalue son utilité clinique en comparant l'âge réel de la colonne vertébrale aux prédictions du modèle.

A novel artificial intelligence framework has been developed to measure human spine aging using MRI, leveraging a deep learning method that analyzes over 18,000 MRI series focused on age-related spine degeneration. The model employs advanced clustering techniques to identify degenerative conditions and evaluates its clinical utility by comparing actual spine age with model predictions.

DelTriC: A Novel Clustering Method with Accurate Outlier

Was this article worth reading? Share it

Competitor Research

Deptho.ai

Dyad