“…Such a variety of modalities describes different clinical aspects of cancer disease and can provide a wide range of complementary bio-markers leading to more accurate diagnosis and more efficient treatment plans. Although there are several works in the current state-of-the-art dealing with the detection, classification and prognostic task taking the aforementioned single modalities individually [18], [19], [20], [21], [22], there are still few works in oncology that aim to fuse these modalities together. Hence, in recent years, researchers focused their efforts on the fusion of these modalities into a single machine learning framework [8], [9], [10], [11], [12].…”