Intelligence Augmentation: Human-centric Healthcare
Author: Pallavi Palparthy, Market Innovation Manager at Philips
There has been an explosion of conversation on artificial intelligence with programs like ChatGPT making frontline news everywhere. We find ourselves on the precipice of the shaping of viewpoints on AI and there in an ongoing debate on the benefit and risks of the adoption of the technology across industries. AI was originally modeled after human intelligence and there have been many advances made in recent years and these advances have led to fear that this technology will dominate and surpass us. Intelligence augmentation (IA) on the other hand is full of promise to turn the tension between AI and human intelligence into a symbiosis.
What is Intelligence Augmentation (IA)?
Intelligence augmentation refers to a partnership between a person and a machine in which both contribute their strengths. It uses AI technologies like machine learning, but instead of replacing humans, it aims to assist them. IA machine learning and the human brain work in combination to empower each other. You will also hear terms such as ‘’intelligence amplification’’, ‘’assistive intelligence’’, ‘’machine-augmented intelligence’’, or ‘’cognitive augmentation’’ used to describe intelligence augmentation. As AI becomes more robust (as seen with ChatGPT) and its impact in art and education amongst other fields, there is a concern that AI technology may negatively affect human life. IA was designed because of fears and concerns about AI as a whole.
The idea that machines can amplify human abilities has existed for many decades. In 1960, J. C. R. Licklider proposed cooperation between people and computers. While technology pioneer Douglas Engelbart (1962) also advocated for IA, he said that by “augmenting human intellect” we mean increasing the capability of a man to approach a complex problem situation, to gain comprehension to suit his particular needs, and to derive solutions to problems.
Intelligence Augmentation in Healthcare:
A technology that focuses on effectively using information technology to augment human intelligence is seen to be very effective in healthcare. In healthcare, human intelligence cannot be replaced, and hence IA specifies systems that augment human intelligence rather than attempt to replace them. Such systems can work to enhance human intelligence and improve human decision-making functions in some way, for instance, a machine learning algorithm can process massive amounts of patient data by searching through a patient’s history, family history, previous records, wearables data, and present this information to medical staff in a way that supports the doctor, who would then use their reasoning to reach a prognosis.
Another example is the tools developed through IA that can be used for image processing, Cambridge consultants have determined potential solutions by combining CT and MRI scans to build 3D images of a person’s body that assists surgeons to perform more precise surgical processes. Technology will never replace knowledge work entirely, but IA can make things easier by saving time and improving efficiency.
There is a reason we find it difficult to let technology like artificial intelligence make decisions about our health and our needs, we believe that it would not consider our unique being or our peculiar mannerisms and circumstances. While many people might experience a headache, ‘’my’’ headache is unique to me. While much has been said about the disruptive potential of technology, we should focus less on the Sci-Fi style application of AI and focus more on opportunities for providing medical care that is faster and leads to better outcomes by accounting for the unique circumstances that apply to an individual.
For instance, in robotic-assisted surgery, augmented intelligence does not replace the skill and expertise that surgeons hone over years of experience and thousands of repetitions. Instead, these skills and experiences are precisely what augmented intelligence can leverage to improve the standard of care. IA can provide real-time feedback during these surgical procedures; this feedback will be based on a vast database of prior procedures. It will then rely on the surgeon’s skill and decision-making ability to create recommendations for this particular patient.
There is nothing “artificial” about a recommendation based on data from a real surgeon in a real surgery, considering the unique circumstances of the patient.
Advantages of Intelligence Augmentation:
· Efficiency: IA provides synergies between machines and medical specialists so that technology can assist the former in providing care faster and more efficiently.
· Reduces Errors: It can help eliminate human errors and fill in gaps that are hard for another human to plug in.
· Better outcomes: When technology and humans work in collaboration, the combination can deliver the best outcomes for both the physician and the patient and improve the physician/patient relationship.
· Decision Support: IA methods combined with analytics can provide additional capabilities to decision support systems for health risk assessment and patient-clinician shared decision-making.
· Quadruple aim: These systems have the power to transform health care and bring us closer to the quadruple aim. At Philips we hold the ‘’quadruple aim’’ as the true north star, in this we aim first to make the diagnosis more effective, we then provide an enhanced patient experience, thirdly we work towards better outcomes for the healthcare practitioner and fourth, we try to lower the total cost of treatment.
Augmented intelligence and machine learning can streamline and give actionable insights long before and after a patient sees a doctor, giving the healthcare industry’s overworked humans precious hours of their time back.
Augmented intelligence follows a five-function cadence that allows it to learn with human influence:
1. Understand: Feed the right data to break down and derive meaning from.
2. Interpret: When new data is inputted, the system then reflects on old data to interpret new data sets.
3. Reason: The system creates “output” or “results” for new data sets.
4. Learn: Humans give feedback on output and the system adjusts accordingly. IA is an intelligence tool with a human feedback loop.
5. Assure: Security and compliance are ensured using blockchain or AI technology
To understand this better, let us look at the process of medical diagnoses. It is based on several aspects, but there are three vital steps — information about symptoms to be treated, the experience of the healthcare professional, and information about past patients with similar symptoms. With these three aspects addressed, there is a strong possibility of a correct diagnosis. It is in the third aspect that augmented intelligence will play a role as it is not humanely possible to compare one patient to a large set of other patients.
For instance, a dermatologist while diagnosing skin lesions can feed a picture of the patient’s lesion into a machine along with their symptoms and past medical history. The intelligent algorithm then uses past data to compare this new data that has been input and comes back with a possible diagnosis and possibly the rationale for the decision. The dermatologist uses this output and applies their experience and training to make a final diagnosis decision. The benefit of using augmented intelligence is that it leads to faster, and more precise diagnosis and is, therefore, beneficial for the entire healthcare ecosystem.
How to attain effective augmented intelligence:
· Cultivate and look for the right talent, i.e., bilinguals with domain and technology expertise.
· Have proper governance over automation and people.
· Think through change management, specifically ensuring a smooth adoption of the new ways of work and upskilling talent.
· Make sure your technology and processes have continuous human input.
· Develop white-box algorithms that can explain why things are not working as they should.
Balancing data privacy and information sharing for secure augmented intelligence:
Data privacy and data sharing are two major concerns for the healthcare industry in adopting the latest technologies. The best way to handle these concerns is by being transparent with patients and consumers and obtaining informed consent from them. This can be the first step in ensuring that they know their data is stored and used for research and analysis.
As a second step, patients should be explicitly told how using their data will improve care, and they should be engaged in a dialog on the topic instead of keeping them in the dark about their data.
Doctors + Technology = Better outcomes
The human brain is an extremely complex and advanced processing unit that can solve problems and still account for the unique circumstances that apply to an individual. No matter how intelligent machines may become they won’t be able to replace humans. So, we have nothing to fear. We only need to prepare. In the foreseeable future, technology will be augmenting our capabilities and allow us to do more in less amount of time. Rather than focusing on whether we will be replaced, we should consider the possibilities of how we can take advantage of augmented intelligence.
…
Interested in making a difference in healthcare? Find an opportunity here!
