arXiv subjects

Salvador E. Ayala-Raggi

Publications and source records attributed to Salvador E. Ayala-Raggi.

2 recordsLinked to original sources

arXiv2026

Accurate Recognition of Pneumonia and COVID-19 by Geometric Shape Normalization of Lung Region using Automatic Landmark Detection and Piecewise Affine Warping

This paper presents an automatic system for recognizing pulmonary diseases in chest X-rays using geometric normalization of the lung region. The method combines three modules: (1) a ResNet-18 landmark detector with coordinate attention that predicts 15 lung-contour landmarks, achieving a mean localization error of 3.61 pixels through an ensemble of four models with test-time augmentation; (2) a geometric normalizer based on Generalized Procrustes Analysis, Delaunay triangulation, and piecewise affine warping to map each lung region to a standardized shape; and (3) a ResNet-18 classifier with transfer learning and SAHS contrast enhancement to classify images as COVID-19, Viral Pneumonia, or Normal. On the COVID-19 Radiography Database, the normalized-image classifier achieved 98.60+/-0.26% accuracy and 98.00% F1-Macro using five-fold cross-validation. Although original images produced slightly higher raw accuracy, Grad-CAM and cropping experiments suggest that this advantage is partly influenced by acquisition artifacts. In contrast, geometrically normalized images outperformed artifact-masked/cropped unaligned images on both the COVID-19 Radiography Database (98.60% vs. 96.24%) and a balanced adult-pediatric mixed dataset including pediatric cases from the Kermany dataset (94.67% vs. 94.17%). These results suggest that anatomical alignment can provide a more controlled and artifact-resistant representation for pulmonary disease recognition.

cs.CV↗

arXiv2025

Is simplicity still possible for a more accurate approximation to the perimeter of the ellipse? or, Using the exponential function to further improve the second Ramanujan's approximation

The perimeter of an ellipse has no exact closed-form expression in terms of elementary functions, and numerous approximations have been proposed since the eighteenth century. Classical formulas by Fagnano, Euler, and Ramanujan, as well as modern refinements such as Cantrell and Koshy methods, aim to reduce the approximation error while maintaining computational simplicity. In this paper, we introduce a new closed-form expression that enhances Ramanujan second formula by dividing it by 1 minus a binomial of two exponential terms resulting in a very stable approximation in a range of b/a between 1 and 1/10000, or even up to a smaller ratio. The resulting approximation remains compact, requiring only four constants, and achieving a remarkable tradeoff between simplicity and accuracy. Across the full eccentricity range of b/a in [0.0001,1], our method attains a maximum relative error of approximately 0.57 ppm with respect to the exact perimeter computed via elliptic integral. Our formula is quasi-exact at the extremes, for the circle b/a=1 and for the degenerate flat ellipse b/a=0. Compared with Cantrell approximation, the proposed method reduces the maximum relative error by a factor of 25 while preserving a short and elegant expression. This makes it one of the simplest yet most accurate closed-form and single-line approximations to the ellipse perimeter currently available in the literature.

math.NA↗