Descripció del projecte
This doctoral project aims to develop new tools based on artificial intelligence, more specifically machine learning and deep learning, for the interpretation and monitoring of rheumatological diseases fundamentally and initially based on imaging techniques based on simple radiology and later adding supporting clinical data.
In a first phase, the study of the interpretation of radiographs of the locomotor system —especially hands, feet, knees, hips and spine— is proposed with the aim of determining the degree of normality and automatically identifying possible pathologies or structural alterations, both benign and malignant.
Subsequently, the development of advanced AI models for the detection of structural changes, whether established as incipient, is foreseen, advancing the capacity for early detection of these lesions and being able to monitor them. This challenge includes the automatic segmentation of small joints (such as those of the hands and feet). In parallel, it is proposed to integrate radiological information with clinical and analytical biomarkers to establish a multimodal analysis framework that allows a more precise characterization of the disease state.
In a third stage, the project focuses on assessing progression and response to treatment using radiographs, with the development of AI models capable of automating the quantification of structural damage and monitoring it. In addition, the aim is to generate predictive systems that, by combining imaging and clinical history, can anticipate patient response to both conventional and biological therapies.
Finally, work will be done on the implementation of AI mechanisms for the automated reading of rheumatic indices, with the aim of facilitating the longitudinal monitoring of patients and contributing to the personalization of therapeutic strategies. In this way, it would be possible to implement in clinical practice all those imaging indices that have been used in clinical trials and in research in general and that have demonstrated their validity in the monitoring of patients but that are not applicable in the day-to-day of conventional consultation, due to time constraints. It would be a step towards excellent care.
This project aims to establish a line of translational research, located at the intersection of rheumatology, radiology and artificial intelligence, with a strong potential impact on both clinical practice and the design of future personalized treatments, prioritizing monitoring of excellence and the identification of individuals at risk of progression; advancing towards precision medicine.
The centers involved in this research project are the Parc Taulí University Hospital, the Parc Taulí Research and Innovation Institute (I3PT) and the Computer Vision Center of the Autonomous University of Barcelona.
