(Semantic) Segmentation of Biomedical Images

Introduction

In digital image processing and computer vision, image segmentation is the process of partitioning a digital image into multiple image segments, also known as image regions or image objects (sets of pixels). Semantic segmentation sums up to recognizing and understanding what is in the image at pixel level: basically, a digital image is partitioned into multiple image regions and a label is assigned to every pixel such that pixels with the same label share certain characteristics. 

In medical imaging, semantic segmentation can recognize different regions that can correspond to: tissue classes, pathologies, organs, other biologically relevant structures, etc. 

It is a challenging task, due to noise, smoke, blur, specular reflections, occlusions.

Our Research

We worked in different scnarios (have a look at the related publications). 

As an example [7], we worked om the assessment of vascular complexity in the lower limbs, as it provides relevant information about peripheral artery occlusive diseases (PAOD), thus fostering improvements both in therapeutic decisions and prognostic estimation. The current clinical practice consists of visually inspecting and evaluating cine-angiograms of the interested region, which is largely operator-dependent. We defined an automatic method for segmenting the vessel tree and compute a quantitative measure, in terms of fractal dimension (FD), of the vascular complexity. The proposed workflow consists of three main steps: (i) conversion of the cine-angiographies to single static images with a broader field of view, (ii) automatic segmentation of the vascular trees, and (iii) calculation and assessment of FD as complexity index. Experimental analyses suggest that extracting the vascular tree from cine-angiography can substantially improve the reliability of visual assessment of vascular complexity in PAOD. Results also reveal the effectiveness of FD in evaluating complex vascular tree structures.

Publications

[1] C. Adornetto, P. Bruno, F. Calimeri, E. De Rose, and G. Greco, "Artificial Intelligence in Medicine: From Imaging to Omics", in CEUR Workshop Proceedings, vol. 3486, pp. 140-145, 2023. 2023 Italia Intelligenza Artificiale - Thematic Workshops, Ital-IA 2023, 29-30 May 2023.

[2] P. Bruno, M. F. Spadea, S. Scaramuzzino, S. De Rosa, C. Indolfi, G. Gargiulo, G. Giugliano, G. Esposito, F. Calimeri, and P. Zaffino, "Assessing vascular complexity of PAOD patients by deep learning-based segmentation and fractal dimension", Neural Computing and Applications, vol. 34, no. 24, pp. 22015-22022, 2022. DOI: 10.1007/s00521-022-07642-2

[3] M. K. Sherwani, A. Marzullo, E. De Momi, and F. Calimeri, "Lesion segmentation in lung CT scans using unsupervised adversarial learning", Medical and Biological Engineering and Computing, vol. 60, no. 11, pp. 3203-3215, 2022. DOI: 10.1007/s11517-022-02651-8

[4] M. Scarfone, P. Bruno, and F. Calimeri, "A Parallelization Approach for Hybrid-AI-based Models: an Application Study for Semantic Segmentation of Medical Images", in CEUR Workshop Proceedings, vol. 3281, pp. 9-20, 2022. 1st International Workshop on HYbrid Models for Coupling Deductive and Inductive ReAsoning and the 29th RCRA Workshop on Experimental Evaluation of Algorithms for Solving Problems with Combinatorial Explosion, HYDRA-RCRA 2022, 5 September 2022.

[5] P. Bruno, F. Calimeri, C. Marte, and M. Manna, "Combining Deep Learning and ASP-Based Models for the Semantic Segmentation of Medical Images", in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12851 LNCS, pp. 95-110, 2021. 5th International Joint Conference on Rules and Reasoning, RuleML+RR 2021, 13-15 September 2021. DOI: 10.1007/978-3-030-91167-6_7

[6] J. F. Lazo, A. Marzullo, S. Moccia, M. Catellani, B. Rosa, F. Calimeri, E. de Momi, and M. de Mathelin, "A lumen segmentation method in ureteroscopy images based on a deep residual U-net architecture", in Proceedings - International Conference on Pattern Recognition, pp. 9203-9210, 2020. 25th International Conference on Pattern Recognition, ICPR 2020, 10-15 January 2021. DOI: 10.1109/ICPR48806.2021.9411924

[7] P. Bruno, P. Zaffino, S. Scaramuzzino, S. De Rosa, C. Indolfi, F. Calimeri, and M. F. Spadea, "Segmentation of vessel tree from cine-angiography images for intraoperative clinical evaluation", in CEUR Workshop Proceedings, vol. 2272, 2018. 2018 RiCeRcA Workshop, RiCeRcA 2018, 22 November 2018.

[8] P. Bruno, P. Zaffino, S. Scaramuzzino, S. De Rosa, C. Indolfi, F. Calimeri, and M. F. Spadea, "Using CNNs for Designing and Implementing an Automatic Vascular Segmentation Method of Biomedical Images", in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11298 LNAI, pp. 60-70, 2018. 17th Conference of the Italian Association for Artificial Intelligence, AI*IA 2018, 20-23 November 2018. DOI: 10.1007/978-3-030-03840-3_5