Final Report Summary - MAMMA (Spatio-Temporal Modeling for Enhanced Automated Detection and Classification of Non-Mass Lesions in Breast MRI)
The emphasis of this project lies in the development and evaluation of an intelligent and robust computer-assisted system for detecting and diagnosing breast lesions that present a non-mass-like enhancement and thus lead to a substantial improvement of the quality of breast MRI postprocessing, reduce the number of missed or misinterpreted cases leading to false-negative diagnosis, and avoid unnecessary biopsies for benign lesions or observation for malignant lesions.
Non-mass-enhancing lesions exhibit a heterogeneous appearance in breast MRI with high variations in kinetic characteristics and typical morphological parameters, and have a lower reported specificity and sensitivity than mass-enhancing lesions. Combinations of morphological and temporal BIRAD descriptors have proven to be insufficient to aid in the automated differential diagnosis of these lesions. We developed novel spatio-temporal descriptors that lead to a substantial improvement in diagnostic accuracy and efficiency and validated them in three specific experiments.
Radiologists can benefit from this system by reduced interobserver variation and improved interpretation of mammograms for the presence or absence of malignant non-mass-like enhancing
lesions. Adding novel algorithms to existing techniques in breast CAD systems will create a flexible toolbox that can be applied with minimal modifications to identifying other type of lesions or monitor response to chemotherapy.
The expected results of this interdisciplinary project will definitely find tremendous clinical applica-tions and impact on the successful treatment disease in the European society, and strongly addresses the overarching goals of the ”2020 Vision for the European Research Area”. Specifically, improving treatment outcome of major diseases such as breast cancer is a research priority in the European Union.
Non-mass-enhancing lesions exhibit a heterogeneous appearance in breast MRI with high variations in kinetic characteristics and typical morphological parameters, and have a lower reported specificity and sensitivity than mass-enhancing lesions. Combinations of morphological and temporal BIRAD descriptors have proven to be insufficient to aid in the automated differential diagnosis of these lesions. We developed novel spatio-temporal descriptors that lead to a substantial improvement in diagnostic accuracy and efficiency and validated them in three specific experiments.
Radiologists can benefit from this system by reduced interobserver variation and improved interpretation of mammograms for the presence or absence of malignant non-mass-like enhancing
lesions. Adding novel algorithms to existing techniques in breast CAD systems will create a flexible toolbox that can be applied with minimal modifications to identifying other type of lesions or monitor response to chemotherapy.
The expected results of this interdisciplinary project will definitely find tremendous clinical applica-tions and impact on the successful treatment disease in the European society, and strongly addresses the overarching goals of the ”2020 Vision for the European Research Area”. Specifically, improving treatment outcome of major diseases such as breast cancer is a research priority in the European Union.