Swarm-Intelligence Strategy for Diagnosis of Endogenous Diseases by Nanobots

An ambitious long-term goal of medicine is to make analyses and deliver drugs selectively at cell level. In particular such "ideal" drugs should be able to travel through the vasculature, reach the intended target at full concentration, and there act selectively on diseased cells and tissues only, without creating undesired side effects. The state of the art in pursuing this goal is represented by \emph{nanoparticles} \cite{Ferrari2010}, which can be roughly defined as the combination of a drug molecule with a suitable vector of nanoscale dimensions. One of the main limitations of nanoparticles is that they do not have an onboard control system. Their ``program'' has to be defined in the drug design phase. A different program means the development of a new nanoparticle. Moreover the lack of onboard control means also that the complexity of the actions to be performed by nanoparticles is quite restricted. The basic idea sustaining this goal is inspired by the immune system. That the immune system is a fantastic machine for surveilling and fighting against exogenous diseases is too well known to deserve discussion. However, it is not perfect. A drawback of the immune system system is the poor recognition of endogenous pathological cells like those responsible for cancer or self-immune diseases. In this context it should be interesting to develop artificial devices (\emph{nanobots}) working as blood white cells but addressed to the recognition and eventually the destruction of endogenous pathological states. A nanobot \cite{Requicha2003} can be defined as any artificial machine with overall size of the order of a few microns or less ($\simeq$ red cells), constituted by nanoscopic components with individual dimensions in the interval $1-10^2$ nm, and able to perform sophisticated functions like navigation, cell recognition, data collection and transmission. To the best of my knowledge no attempt toward the definition of an auxiliary immune system is however known. Of course, this goal is extremely ambitious and is expected to require decades. Nonetheless, sketching a scenario is not a mere speculation, but rather is useful to identify the nature of problems posed by the development of nanobots for real applications. One of the possible reasons for the ineffectiveness of the immune system in detecting malignant cells is that the common features associated with cancer genesis and growth (hyperthermia, hypoxia and acidification) are the same as those characteristic of muscle under physical exercise. My program is thus that of \emph{endowing the immune system with an artificial surveillance system devoted to detecting the simultaneous occurrence of hyperthermia, hypoxia and excessive acidity due to localized cancerous states without confusing it with the similar conditions produced under physiological conditions}. Considering the diagnosis of cancer as a crime detection, the search of the perpetrator is not addressed to the identification of the smoking gun but rather to a frame evidence resulting from the simultaneous occurrence of three events (hyperthermia, hypoxia and excessive acidity) and their persistence in time. The aim of this thesis is twofold. On one side I want to describe an architecture of nanobots which is not only able to embed sophisticated functions, but also suitable for being manufactured by processes compatible with today and likely tomorrow semiconductor industries. On the other side, I want also to analyze and develop the algorithms needed for successfully tackling the nanobot tasks. In particular, since the nanobots would be devices with very limited computational resources and the interaction between them could only happen locally, the algorithms to control them could be based on swarm intelligence; i.e. it could be inspired by the collective behavior of social-insect colonies and other animal societies.