Who we are
Board of Recommendation
How to Become a Member
Members’ Activities Calendar
What we do
Policy and Advocacy
Exchange the world
Introduction to IFMSA Exchanges
List of Participating Countries
Research Projects Database
Medical Students International
You are here:
Artificial Intelligence developments in radiology: radiomics machine-learning classification of skeletal cartilaginous tumors
Italy (SISM) - University of Milan, Milan
Department of Diagnostic and Interventional Radiology
Dott. Salvatore Gitto, Dott. Carmelo Messina
Dott. Salvatore Gitto, Dott. Carmelo Messina
Type of Research Project
- Clinical Project without Laboratory work
What is the background of the project?
Bony cartilaginous tumors are frequently encountered in clinical practice and include benign entities and malignant chondrosarcomas. Reliable identification and grading are crucial for the clinical outcome, because management ranges from watchful waiting (for benign neoplasms) to minimally invasive-to-aggressive surgery (for low-to-high grade chondrosarcoma). The diagnosis relies on clinical presentation, imaging and histology. However, low reliability among expert pathologists as well as among expert radiologists has been found in correctly classifying cartilaginous tumors and defining tumor grade because of overlapping histological and imaging features. Thus, there is a need for more accurate diagnostic tools. Radiomics includes extraction, analysis and interpretation of quantitative features from medical images. In particular, texture analysis is an emerging radiomics method for quantification of tumor heterogeneity, a key feature of malignancy that is hard to capture using conventional imaging or sampling biopsies. Radiomics can be combined with machine learning algorithms to identify the best combination of quantitative features and create predictive models for the diagnosis of interest.
What is the aim of the project?
The aim is to assess the diagnostic accuracy of machine learning for discrimination among cartilaginous bone tumors based on radiomics parameters extracted from medical imaging.
What techniques and methods are used?
This study will retrospectively enroll patients with histologically-proven cartilaginous bone tumor and imaging available for quantitative analysis, including computed tomography (CT) or magnetic resonance imaging (MRI). Quantitative imaging features will be extracted and analyzed using dedicated softwares. Machine-learning models will be created on the basis of the extracted features and then tested in order to classify tumors and predict tumor grade. The diagnostic performance of the radiomics method will be compared with that of an experienced radiologist.
What is the role of the student?
- The student will observe the practical experiments but will be highly involved in the analysis of the results
- The tasks of the student will be performed on his/her own
- The tasks will be done under supervision
What are the tasks expected to be accomplished by the student?
Students will be expected to: - understand the rationale behind the clinical use of imaging in the diagnosis of bone tumors - become familiar with the basic concepts of imaging modalities used to assess bone tumors - become familiar with the basics of research methodology, including collection and analysis of data, in the field of radiomics and quantitative imaging - have a major role in the presentation of preliminary data at conference meetings - participate in the drafting of the final paper
Will there be any theoretical teaching provided (preliminary readings, lectures, courses, seminars etc)
The student will familiarize themselves with the imaging methods available for the diagnosis of bone cartilaginous tumors and research methodology used in the field of radiomics. This will be done by means of preliminary readings, looking at cases under supervision of radiologists and on-site lectures.
What is expected from the student at the end of the research exchange? What will be the general outcome of the student?
- The student will prepare a poster - The student will prepare an abstract - The student’s name will be mentioned in a future publication - The student will have the opportunity to present the results together with the supervisor at a conference
What skills are required of the student? Is there any special knowledge or a certain level of studies needed?
Initiative, curiosity, critical thinking, adaptability. Basic knowledge of medical imaging techniques such as CT and MRI.
Are there any legal limitations in the student’s involvement
Type of students accepted
This project accepts: - Medical students - Graduated students (less than 6 months)
- Fritz et al. Magnetic Resonance Imaging-Based Grading of Cartilaginous Bone Tumors: Added Value of Quantitative Texture Analysis. Invest Radiol 2018; 53:663-672.
© 2015 - IFMSA.org - Developed by web agency