Biomedical Imaging Laboratory

Objectives:

The activity concerns the analysis, design, and implementation of machine learning methods for multimedia applications involving biomedical images and image sequences.

Main achievements:

Microarray images: we proposed an automatic approach to gridding (also known as addressing or spot finding) in microarray images, based on the Orientation Matching and the Discrete Fourier Transforms [1].


Dermoscopic images: we proposed an approach to the characterization of uncertain lesions, in order to detect characteristic profiles of benign and malignant lesions [2]. Based on suitable multi-value descriptors (scalar, interval, and hystogram data) extracted by dermoscopic images, it consists in selecting through discriminant analysis the most discriminating features and detecting through dynamic clustering the characteristic profiles.

Moreover, we proposed the SDI algorithm for the automatic segmentation of skin lesions in dermoscopic images [6], extensively tested on the lesion segmentation dataset made available for the ISIC 2017 challenge on Skin Lesion Analysis Towards Melanoma Detection.

Ongoing researches:

ESC images: segmentation and classification of microscopy images aimed at identifying specific factors that control the differentiation of Embryonic Stem Cells via large scale screenings [3][5].


HeLa cell image sequences: segmentation, tracking, and lineage of HeLa cells from phase contrast microscopy time-lapse data [4].


Main collaborations:

- Computer Vision and Pattern Recognition Laboratory (CVPRLab), University of Naples Parthenope, Naples, Italy

- Dept. of Dermatology, Second University of Naples, Naples, Italy

- Dept. of Political Science “J. Monnet”, Second University of Naples, Caserta, Italy

- Inst. of Genetics and Biophysics, National Research Council, Naples, Italy

Publications on Biomedical Imaging:

[6] M. R. Guarracino and L. Maddalena, Segmenting Dermoscopic Images, arXiv:1703.03186, 2017.

[5] L. Casalino, P. D’Ambra, M. R. Guarracino, A. Irpino, L. Maddalena, F. Maiorano, G. Minchiotti, E. J. Patriarca, Image Analysis and Classification for High-Throughput Screening of Embryonic Stem Cells, in V. Zazzu, M.B. Ferraro, and M.R. Guarracino (Eds.),Mathematical Models in Biology, Springer, pp 17-31, 2015.

[4] M. Sangiovanni, L. Maddalena, M. Guarracino, Following the Changes: HeLa cells Lineage from Phase Contrast Microscopy Time-Lapse Data in “6th International Workshop on Data Analysis Methods for Software Systems”, pag. 45, dic. 2014.

[3] Image Segmentation and Classification for High-Throughput Screening of Microscopy Imagery, Conference Horizon 2020@DIITET, CNR, May 26-27, 2014.

[2] V. Cozza, M.R. Guarracino, L. Maddalena, A. Baroni, Dynamic Clustering Detection through Multi-valued Descriptors of Dermoscopic Images, Statistics in Medicine, John Wiley & Sons, Ltd., Vol. 30, Issue 20, pagg. 2536–2550, DOI: 10.1002/sim.4285, 2011. [Impact Factor: 2.328]

[1] L. Maddalena, A. Petrosino, Metodi per l'Analisi di Immagini da Microarray, Tutorial “Metodi e Strumenti per l’analisi dei Dati di Espressione Genica”, Naples, December 2007.

Last update: March 10, 2017