This paper considers the advantage of knowledgeaided direct targeting strategy (DTS) over systemic targeting for tumor homing in complex human vasculature. Tumor location in the body can be estimated by closely observing the tumortriggered bio-physical gradients in its vicinity, helping drugloaded magnetic nanoswimmers to accumulate at the disease location. These nanoswimmers are assembled by magnetic nanoparticles (MNPs), which act as contrast agents increasing the diagnostic capability of different medical imaging techniques. We propose a novel DTS inspired by the iterative gradient descent (GD) with momentum optimization for tumor targeting amplification. We show by computational experiments that the MNPs accumulate at the disease location at a faster rate and the probability of tumor detection is higher for the proposed knowledge-aided DTS as compared to the knowledgeless systemic targeting. We believe that our work for tumor targeting amplification based on such nanosystem will open new horizons in the field of diagnosing tumor at its early stage.