| KEYWORD |
DAUIN - GR-09 - GRAphics and INtelligent Systems - GRAINS
Reti neurali profonde per la segmentazione automatica di tumori cerebrali in risonanza magnetica
Parole chiave DEEP NEURAL NETWORKS, MEDICAL IMAGING
Riferimenti LIA MORRA
Gruppi di ricerca DAUIN - GR-09 - GRAphics and INtelligent Systems - GRAINS
Descrizione Brain tumor segmentation from neuroimaging modalities is a critical step towards improving disease diagnosis, treatment planning and post-treatment monitoring. Segmentation methods are mainly grouped into three categories: manual, semi-automatic and fully automatic. As manual segmentation of brain tumors is time-consuming and subject to rater variability, the interest for automatic and semi-automatic segmentation of brain tumors has received a great impulse over the last decades. While many efforts have been done to improve the segmentation of the pre-operative scans, artificial intelligence has not obtained comparable results for post-operative images.
The aim of this study is to make a further step in this field by the application of deep learning technologies to post-operative neuroimaging. The main focus is the creation of a reliable algorithm for the semi-automatic analysis of MR images, including both pre-operative and post-operative scans obtained with standardized technique and timing. This achievement would have relevant implications for the post-operative management of patients and for the possible prediction of recurrent disease features, including timing and site.
SUGGESTED READINGS:
1. Baid, U. et al. The RSNA-ASNR-MICCAI BraTS 2021 Benchmark on Brain Tumor Segmentation and Radiogenomic Classification. https://arxiv.org/abs/2107.02314
2. Eijgelaar, R. S. et al. Robust Deep Learning–based Segmentation of Glioblastoma on Routine Clinical MRI Scans Using Sparsified Training. https://pubs.rsna.org/doi/pdf/10.1148/ryai.2020190103
3. EH, F., AL, S., EM, B., O, S. & I, R. Glioblastoma Segmentation: Comparison of Three Different Software Packages. PLoS One 11, (2016).
4. Ermiş, E. et al. Fully automated brain resection cavity delineation for radiation target volume definition in glioblastoma patients using deep learning. https://ro-journal.biomedcentral.com/track/pdf/10.1186/s13014-020-01553-z.pdf
Conoscenze richieste deep learning, machine learning, python
Scadenza validita proposta 25/11/2023
PROPONI LA TUA CANDIDATURA