Artificial intelligence is able to see things that doctors miss, with a very high level of detail and accuracy. To the point that it is possible to exploit it to photograph and analyze the movements of patients suffering from degenerative diseases: the patient wears sensors and moves, the software reconstructs his image into an avatar and allows doctors to understand how to intervene to improve the symptoms or evaluate the effectiveness of any drugs, speeding up research. A team of British researchers – including Italian scientists working across the Channel – is convinced of this and is betting on revolutionizing research into rare diseases in this way.
Building new disease markers
Two studies published on the pages of Nature Medicine, dedicated to the promises of wearable devices and artificial intelligence in Duchenne’s disease and Friedreich’s ataxia, degenerative pathologies that compromise (among other things) the patients’ ability to move, to the point of making it impossible to walk. The idea of the researchers was to understand if it is possible to use wearable sensors to extract as much information as possible on the movement of patients. The logic is no different from that of the motion capture systems used in video games: the person moves, the sensors record the movements which are then processed by an artificial intelligence, capable of extrapolating data that can be used as a real disease marker . But that’s not all: these same data can also be used to develop forecasts on the evolution of the disease, in more detail than traditional clinical evaluations. In essence, it is a question of studying the digital behavior of patients, the researchers explain, by exploiting the ability of wereables and AI to perceive characteristics that are not always identifiable by the medical eye.
Big data in medicine: it’s time to invest
by Annalisa Bonfranceschi
New possibilities for research
The predictive capabilities of these systems are so high that, in the case of Friedreich’s ataxia, they even manage to identify the expression level of the gene affected in the disease (that of the protein frataxin, FXN). But more generally, the true potential of analyzes like these is to identify so-called kinematic markers potentially useful for evaluating responses to therapies, which promise objectivity and sensitivity from them. Overcoming the limitations of clinician assessments and traditional scales, which sometimes struggle to identify minimal changes in disease, write the authors.
Breast cancer, artificial intelligence increases the sensitivity of ultrasound
by Mara Magistroni
More generally, the availability of new disease markers is all the more useful considering that we are talking about rare diseases, where the scarcity of patients is one of the limits in the search for new treatments, as the authors acknowledge. “Research on rare diseases can be considerably more expensive and logistically demanding – comments Valeria Ricotti of the Great Ormond Street Institute of Child Health of University College London, one of the authors of both papers in Nature Medicine – this means that patients they are missing out on new treatments. Increasing the efficiency of clinical trials gives us hope that we can successfully test many more treatments.”