The primary objective of the BIOMEDE-IA project (2020-2023) is to predict by machine learning methods the main genomic mutations of patients with diffuse intrinsic pontine gliomas from their clinical and imaging (multi-parametric MRI) features, to manage the cases for which biopsy is not available. In a second step, by jointly analyzing the available data set, we will seek to define the characteristics of subjects with a survival of more than two years. This project, carried out in collaboration with Gustave-Roussy, Necker Hospital and Neurospin, is based on data acquired by clinical trials BIOMEDE (2014-2019) and BIOMEDE-2 (2020-2023).

LIFEx is a user-friendly image processing software, making it possible for colleagues without any programming skills to perform radiomic studies using any type of medical images, including PET, MR, CT, SPECT or US images. Software development is performed by Christophe Nioche, a research engineer in LITO, and the software is freely available to the scientific community. The software is constantly evolving to account for the suggestions and contributions of users, with whom we maintain tight interactions. All radiomic features available through this software do conform the recommendations of the international IBSI consortium we belong to.

LIFEx currently has more than 3000 users across the world. Its initial version has been described in a 2018 Cancer Research paper. All you may want to know about LIFEx (download, latest version, FAQ, tutorials) can be found on a dedicated web site.

Photo Dynamic Therapy (PDT) is a localized treatment using a laser beam focused on the tumor. The PDT is based on the activation by specific wavelength light of a photosensitizer (PS) localized preferentially in tumor cells. This generates oxygenated species toxic to the cells internalizing PS. PS is non mutagenic and pharmacologically inactive without lighting. We have shown that PDT is effective on retinoblastomas developing in the eyes of transgenic mice.  Our project aims at evaluating PDT on a model developing tumours in the vitreous or retina in rats. This PDT approach will be compared to a more conventional radiotherapy treatment. We will also evaluate this treatment with uveal melanoma and compare its efficiency with that of radiotherapy which is a gold standard for this pathology.

The PRECISION PREDICT project (2020-2022) is led by Institut Curie (Thoracic Oncology Department, Data Department, and LITO) and funded as part of the Health Data Hub call related to the "Improvement of medical diagnosis through the use of Artificial Intelligence".  Together with 8 other Cancer Centers in France, the goal is to create a large database of clinical and radiological images (PET/CT and CT scans) data in patients with lung cancer associated with an EGFR mutation and treated with targeted therapy (Tyrosine Kinase Inhibitors). Using this database, our aim is to better understand the observed variety of response to the treatment among patients, and to try predict, for each individual patient, whether he/she will benefit or not from the treatment and for how long. 

VOCALE is a project (2018-2022) dedicated to motion analysis of vocal folds using dynamic translaryngeal ultrasound. It is driven by the laboratory in collaboration with surgical departments and the "Laboratoire d’Imagerie Biomédicale". We developed a dedicated image analysis software and demonstrated its usefulness to detect and quantify vocal fold paralysis on 100 patients with voice disorders after thyroid surgery. A multi-centric clinical trial is under way to validate the use of ultrasound as first-line exam and reduce the number of nasofibroscopies. In this trial, we are combining image analysis methods with deep learning to improve the performance of our detector of vocal fold paralysis associated with a recurrent nerve injury.


We are part of the H2020-MSCA-ITN-2017 HYBRID (Healthcare Yearns for Bright Researchers for Imaging Data, 2017-2021) project in collaboration with 10 partners from Germany, Austria, United Kingdom, the Netherlands, Danemark, and Belgium. The HYBRID project supports 15 PhD students that are all working towards more powerful in vivo molecular imaging, especially involving PET/MR, for personnalised medicine, using the most advanced technologies and processing approaches, including dynamic imaging and artificial intelligence approaches. To know more about the HYBRID project, please go to the dedicated web site.

The aim of the PANACEE project (2020-2023) is to develop methods and a tool that will make it possible, for a patient with non small cell lung cancer described by his or her clinical, biological, histological or medical images characteristics (radiomics), to identify a small group of patients with very similar characteristics, in a reference database consisting of patients already treated for the same pathology. The medical history of these "twins" will allow doctors to access valuable information to suggest the therapeutic strategy to be adopted for the new subject.This project is funded by the Janssen Horizon Foundation.


The TIPIT project (2020-2023), who involves U900 Inserm - Institut Curie - PSL (Head : Emmanuel Barillot), the Department of thoracic oncology of Institut Curie (Head : Nicolas Girard) and our lab is funded by the ARC fondation, as part of the SIGNIT call,  for 3 years. The project will consist in gathering a database of genomic, radiomic, clinical and pathological data in 200 patients with non small cell lung cancer to get a comprehensive view of each tumor and of its microenvironment. By mining these data, we will characterize each tumor using a multiscale approach, and will try to predict their response to immunotherapy using artificial intelligence methods. The goal is to get a prediction that would be accurate enough to be clinically useful for patient management.  

Positron Emission Tomography is well-established in the diagnosis and treatment monitoring of Hodgkin‘s lymphoma (HL).  HL is a type of tumour that can be characterized by its sugar consumption. Thus, PET enables early monitoring of treatment response, that is shows if the lymphoma shrinks or grows, reflecting whether the therapy is working or not working . So far, only very basic information of the acquired PET images is used. Since the biology of HL is linked to its metabolism, we aim to analyze PET images in more detail by using artificial intelligence algorithms. By doing so, we seek to identify those patients specifically who suffer from aggressive HL and who are in need of more intense treatment. The same algorithms can hopefully be used also to identify patients with a favourable prognosis and who require less intense treatments with fewer side effects. This project (2020-2023) has been funded through the EraCoSysMed 3rd joint translational call (ANR-19-SYME-0005-03 contract).