About the project
Acute lymphoblastic leukaemia (ALL) is the most common form of childhood leukaemia. Thanks to continuously improved treatment strategies, the survival rates of children and adolescents with ALL are rising steadily and are now over 90%. As a result of this positive development, therapy-associated and disease-related adverse events are increasingly coming to the fore. These events are problems/complications that can be caused by the disease itself or by the therapy. These complications can have a major impact on children’s health and can increase the risk of illness and mortality. In children with ALL, therapy-associated and disease-related adverse events include pulmonary complications, such as pneumonia caused by fungal infections, or effects on bone health like reduction of bone mineral density.
The most important facts at a glance
Main objective of the project:
The main objective of the project is a standardised and structured evaluation of serious adverse events of paediatric ALL on a national level. The focus is on pulmonary complications, in particular fungal infections, and on analysing treatment-associated changes in bone health during the course of ALL therapy.
Hypothesis 1: The use of AI-based image processing models enables early detection and characterisation of invasive fungal infections in paediatric ALL patients and reliable differentiation from other pulmonary complications.
Hypothesis 2: Image-based risk factors, parameters and biomarkers can be extracted from chest CT images and used together with clinical data and structured findings in a multimodal deep learning model to predict treatment outcome and progression of pulmonary complications in paediatric ALL patients.
Hypothesis 3: ALL therapy influences the bone health of paediatric ALL patients, which can be objectively recorded, longitudinally quantified and related to a specific therapy regimen or therapy phases using AI-based analysis.
- Due to an estimated Germany-wide case number of approx. 400 paediatric patients with pulmonary fungal infection and available CT thorax examinations, the participation of all possible locations is very important.
- Cooperation between the paediatric oncology and paediatric radiology/radiology departments of the participating sites, including the establishment of the regulatory framework (ethics & data protection)
The RACOON infrastructure is used to merge CT thorax data from routine clinical practice from the vast majority of German university hospitals with comprehensive clinical parameters from the ALL-BFM study group to create a nationally unique multimodal ALL dataset. The resulting data set is analysed using AI-based analysis methods.