The development of microfluidic has made possible to duplicate the functionality of living beings by organising living cells on a chip and subjecting them to conditions that replicate the one of those cells in an organ. Once the functionality (and structure) is replicated it becomes possible to change the condition
(like varying nutrients, altering oxygen perfusion, introducing bacteria and drugs..). By observing the reaction of the cells, and the change in functionality, it becomes possible to simulate in vitro the behaviour of an organ under different conditions. The progress in this area has been towards the replication of a broader set of organs (we now have chips for lung, liver, kidney, gut, skin, heart, pancreas) and to include more functionalities. For simulating some organs’ functionality cells have to be organised in very specific structures, mimicking the one in the real organ and for this bio 3D-printing is used.
Lately researchers have started to connect various organs on a chip to create more complex systems with the objective of creating a “body-on-a-chip” that can represent the complex interactions among the various body systems. Notice that this is not creating a “body”, only a lab system to mimic specific functionality (besides a brain on a chip is still in the science fiction domain, although neurones are studied in vitro and have been integrated in bio-chips).
A first, obvious, use of this technology is for drug discovery/ testing. As a matter of fact the process leading to the commercialisation of a drug is a lengthy and costly one (a drug may cost 2 billion $ taking into account all the dead alleys that have to be discarded). Microfluidic evolution along with data processing is expected to shorten the time to market and decrease significantly the number of dead alleys by the end of this decade.
A further expected evolution is the support to personalised cure, where a body-on-a-chip can be created using the patient cells to test a therapy in vitro. Here again, as previously mentioned, the availability of a personal digital twin may become important since part of the testing could be performed in the cyberspace.
Finally, by the end of this decade, early next one, technology of organ on a chip can become so effective that (in some cases, like pancreas) the chip can be used to replace/flank the natural organ.
5. artificial intelligence
Artificial Intelligence is already used in medical field -applied AI-, particularly as support to diagnoses and prediction (data analytics). Predictive technology has become an important sector of AI in healthcare, looking at improving clinical decision making, managing and leveraging on EHR for risk assessment, monitoring data generated by wearable to raise red flags and augmenting medical devices. This latter will prove particularly important as we will be shifting in this decade towards more and more home care, preventative and personalised medicine.
Interestingly it is also starting to be used in detecting patterns pointing to processes resulting in better outcome (shorter hospitalisation, prompter recovery, fewer side effects/complications). The point is to use AI to find questions, not to find “answers”! The fact is that there are so many data, and data streams available in a hospital environment or in a medical practice that it is possible to apply data analytics to derive questions like “Why are there patterns leading certain patients experiencing better outcome than others?”
In this decade we can expect increased pervasiveness and use of AI in healthcare, as shown in the market forecast by Markets and Markets along the whole value chain:
- new drug design,
- assisted diagnoses, data rendering, automatic EHR updating,
- early diagnostic based on multiple data streams (ALS, cancer, dementia,…)
- assisted surgery,
- therapy identification and monitoring,
- assessing/predicting risks in specific population and at individual level,
- home healthcare, rehab,
- tele-care through virtual doctor,
- robotic/chatbot assisted healthcare, including natural language interaction, virtual nurses,
- doctor continuous education and medicine education for future doctors,
- identification of pathogen, of incipient epidemics, noxious substances,
- genotype to phenotype correlation, diseases related to genetic predisposition and ambient factors
By the end of this decade AI will be pervasive in the whole healthcare area, giving a fundamental contribution to personalisation, monitoring and cost reduction. It can be expected that AI will be embedded in services and medical devices, to be provided both at a centralised level and at a local level. In particular we can expect the embedding of AI in personal digital twins and to have this local intelligence growing over time and cooperating with centralised intelligence. PDTs are also expected to play the role of advisors, steering towards healthier behaviour the physical twin as well as raising the physical twin awareness on potential risks.
The growing pervasiveness of AI in this sector raises questions on privacy, accountability, transparency as well as on the digital divide that is created among the ones that will have access to this support and those that will not be able / allowed to access it. Moreover, there are deep ethical questions in the use of AI. As mentioned cost is an important factor, often a driving deciding factor in the selection of cures. Cost aspects are necessarily part of the AI systems supporting healthcare. Who will be in charge of creating the ethical framework and of checking that it is being followed (this clearly connects to the transparency aspect)?