Atherosclerosis induces abnormal blood-flow patterns in coronary arteries mainly because of the irregular accumulation of fibrofatty plaque on the artery walls. Morphological alterations of the wall-surface attributes downgrade vascular elasticity, which compromises the normal blood-pressure gradient behavior by sporadically interfering with the effective-flow cross-section adjustment. An iron-oxide booster has been recently studied as a potential nano-particle treatment to regulate an elevated blood-pressure condition. Taguchi-type multi-factorial experimentation rapidly generates small and dense datasets in order to expedite arterial flow screening/optimization predictions. The optimal blood pressure-drop performance is investigated against four vital controlling factors. We show that tracking down inherently complex blood-flow phenomena often entails the elucidation of non-linear and messy data structures. Translating such data demands robust and agile techniques to decipher governing relationships while guarding against spurious effects from uncertainty asymmetry. We also show that by using distribution-free profiling, we may synchronously accomplish the screening and optimization tasks more accurately in comparison to other competing techniques. Illustrating our technique on a chemical engineering paradigm, we found that out of the four investigated factors only the blood behavior index to be strongly influential. A blood behavior index setting of 0.5, which is below the normal physiological limit, minimizes the blood pressure drop at an optimal value of 385 Pa/m. The proposed methodology demonstrates how the required sampling size may be further reduced, thus making the study even more economically and practically efficient. This was shown to be achieved without relinquishing important information about the dominant phenomena, hence rendering our solution to be also lean and agile. (C) 2019 Elsevier Ltd. All rights reserved.