The doctoral dissertation in the field of Medical Image Analysis will be examined in Rome, Italy, and it will be streamed online.
What is the topic of your doctoral research? Why is it important to study the topic?
My research focuses on the automated segmentation of medical images acquired using Magnetic Resonance Imaging (MRI). When performed manually, we risk introducing a bias based on the operator. The automated deep learning method presented in my work provides an accurate and efficient alternative to manual segmentation. Furthermore, this is demonstrated in a study looking for anatomical biomarkers of epilepsy based on the position of the hippocampus.
What are the key findings or observations of your doctoral research?
Beside developing a novel image segmentation method for mouse and rat brain MRI, my research shows that previous automated methods, based on aligning the image with one or more atlases, introduce a bias for healthy anatomy. Furthermore, my research shows that it is possible to discriminate between healthy and epileptic rats based purely on anatomical information, using a method potentially translatable to humans.
How can the results of your doctoral research be utilised in practice?
The method presented can be used to accurately and rapidly outline different anatomical structures in a large amount of brain scans in a short period of time, reducing the biases owed to human error and older segmentation methods. This can be of interest for research institutions and private companies dealing with large amounts of MRI scans.
What are the key research methods and materials used in your doctoral research?
The animal data utilized in my research was acquired entirely in past studies, from UEF and from Charles River Laboratories. The remainder of my data analysis work was performed using UEF workstations. The manual work towards annotating the MRI scans to train and test the presented deep neural network was performed in CR, UEF, and in small part by me.
The doctoral dissertation of Riccardo De Feo, MSc, entitled Convolutional neural networks for the segmentation of small rodent brain MRI is a joint project between La Sapienza University of Rome and the Faculty of Health Sciences at the University of Eastern Finland where Professor Jussi Tohka is the main supervisor. The doctoral dissertation will be examined at La Sapienza University of Rome by an examination committee formed by Professors Massimiliano Papi of Universitá Cattolica del Sacro Cuore, Italy, Graziella Messina of the University of Milan, Italy, Emanuela Mercenaro of the University of Genoa, Italy, and Fristed Eskildsen of Aarhus University, Denmark.
