February 2022

PhD defense ESR 11 Fadila Zerka

On February 9 2022 Fadila Zerka succesfully defended her PhD thesis at Maastricht University. Topic of her thesis was: Distributed learning for optimal Radiomics knowledge.  This thesis evaluated the current state of research in the field of radiomics and presented an up-to-date overview of distributed learning applications in health care and their limitations. It also presented a new infrastructure to address the limitations of the existing distributed learning solution. On behalf of the whole consortium we would like to say congratulations! Good luck in the future dr. Zerka and thank you for your the nice collaboration. 

November 2021

A non-invasive, automated diagnosis of Menière’s disease using radiomics and machine learning on conventional magnetic resonance imaging: A multicentric, case-controlled feasibility study

Menière’s disease (MD) is a multifactorial condition of the inner ear characterized by recurrent episodes of vertigo and fluctuating aural symptoms like hearing loss, aural fullness, and tinnitus. The clinical diagnosis of MD, however, is complicated due to the diverse clinical presentation of the disease, symptom overlap with other etiologies, and the lack of specific biomarkers. Therefore, new imaging techniques are under investigation as a MD diagnostic. A study by PREDICT ESR Akshayaa Vaidjanathan and others investigated the feasibility of a new image analysis technique (radiomics) on conventional MRI for the computer-aided diagnosis of Menière’s disease. The multi-layer perceptron classification model yielded a precise, high-diagnostic performance in identifying patients with Menière’s disease based on radiomic features extracted from conventional T2-weighted MRI scans. In the future, radiomics might serve as a fast and noninvasive decision support system, next to clinical evaluation in the diagnosis of Menière’s disease.

An article detailing this work was published in November 2021 in Magnetic Resonance Imaging.

https://link.springer.com/article/10.1007/s11547-021-01425-w

August 2021

PhD defense ESR 5 Marta de Silva Ferreira

On the 24th of August Marta de Silva Ferreira succesfully defended her PhD thesis at Hospitalier Universitaire Liege. During her Phd she extracted radiomic features from different medical images of the cervix (PET, CT and MR) and develop models for predicting treatment outcome of cervical cancer patients.On behalf of the whole consortium we would like to say congratulations! Good luck in the future dr. Ferreira and thank you for your the nice collaboration. 

June 2021

PhD defense ESR 9 Ronrick Da-ano

On June 25th Ronrick Da-ano succesfully defended his PhD thesis at the Université de Bretagne Occidentale. Topic of his research was: Harmonization strategies for multicenter radiomics studies today. Ronrick is the first of the 14 PREDICT ESR's to recieve his doctorate. On behalf of the whole consortium we would like to say congratulations! Good luck in the future Dr. Da-ano and thank you for your the nice collaboration. 

March 2021

Interactive contouring through contextual deep learning

Many clinical procedures rely on accurate contouring of anatomical structures in medical images. For example, image segmentation is extensively used in radiotherapy planning to identify healthy and cancerous regions. Accurate segmentation of both the tumor and the healthy tissue in the image, is essential to maximize the dose of radiation delivered to the tumor while minimizing the dose delivered to healthy tissues. A study by PREDICT ESR Michael Trimpl and colleagues  investigated a deep learning approach that enables three-dimensional (3D) segmentation of an arbitrary structure of interest given a user provided two-dimensional (2D) contour for context. The segmentation performance for seen and unseen structures improved when the training set was expanded by addition of structures previously excluded from the training set. Training a contextual deep learning model on a diverse set of structures increases the segmentation performance for the structures in the training set, but importantly enables the model to generalize and make predictions even for unseen structures that were not represented in the training set. This shows that user-provided context can be incorporated into deep learning contouring to facilitate semi-automatic segmentation of CT images for any given structure. Such an approach can enable faster de-novo contouring in clinical practice.

An article detailing this work was published in March 2021 in Medical Physics.

https://aapm.onlinelibrary.wiley.com/doi/full/10.1002/mp.14852