Deep Learning in Healthcare Part 1: Opportunities & Risks


The application of artificial intelligence and deep learning in healthcare and medicine is often quoted to grow tenfold in 5 years, from algorithms that learn to recognise complex patterns within rich medical data, to analysing real world evidence for personalised medicine, to discovering the sequence specificities of DNA-binding proteins and how it can aid genome diagnostics.

As part of our ongoing speaker Q&A series, we asked deep learning and healthcare experts for their predictions for the field over the next 5 years, the risks involved with AI integration, areas for disruption and more.

Diogo Moitinho de Almeida is a data scientist, software engineer, and hacker. He joined Enlitic as Senior Data Scientist in 2014, where he develops deep learning algorithms for medical diagnosis.

Sobia Hamid holds a PhD in Epigenetics from the University of Cambridge, and is Founder of Data Insights Cambridge, a nonprofit community of over 800 data science practitioners.

Olexandr Isayev is a Research Scientist at UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, where his research focuses on making sense of chemical data with molecular modelling and machine learning.

Vinay Kumar is a researcher in nanotech and co-Founder and CEO of, a developer platform for artificial intelligence with deep learning tools in language, vision, text, speech, dialogues and reasoning.

Michael Nova is Chief Innovation Officer and a co-Founder of Pathway Genomics, as well as the inventor of the Pathway-IBM Watson deep machine learning and AI mobile application Panorama/OME.

What areas of healthcare have the biggest potential for disruption by AI?
Sobia: Results from early applications and demand in the market demonstrate that AI has huge potential to transform medicine and healthcare, and I believe we will see a significant impact within the next 3 years. Key areas showing the greatest potential are automated intelligent medical diagnosis, therapeutic recommendations and healthcare management. New AI technologies are already making inroads into helping to improve efficiencies, accuracy, and cost-effectiveness of medical practice, and will ultimately help us to move closer towards tackling challenging common syndromes, chronic illnesses and rare diseases.

Diogo: Medicine is inherently a data problem, so all of it is up for disruption. A lot of experts have commented that deep learning is missing a killer app, where all we've gotten so far is a better image search, and we think medicine will be AI's killer app. We are particularly excited about diagnostics, treatment planning, and population health because this maximizes the potential worldwide impact.

Vinay: I believe AI adoption will happen in three key stages - Automation, Assistance, Management. Any automation is an easy sell from both Organization and Operational perspective. Within Health Care, 'Primary Health Care' is the first segment that has huge opportunities where value can be added. A smart AI based solution would reduce the burden on doctors and other health care professionals. Of course the maximum impact would be in areas where doctors are scarce or just not available! More generally, any doctor would benefit from having a second opinion available from a peer or a specialist. With machines as 'Digital Assistants', a doctor can get quick opinion on multiple complexities of situations. More advanced forms of assistance can from multi functional robots - for example, a robot that can extract the information from CT scans and search for a solution in text knowledge sources like - research papers, previous reports etc can help doctors greatly. These multi-functional robots can not only change 'Primary Health Care' but also transform the major pre & post diagnosis assistance.

Michael: AI will enable better diagnoses and personalized medical recommendations. AI- enabled autonomous health scans provide the best diagnostics equally to the poorest and wealthiest on Earth.

Olexandr: As machine learning is applied to big medical data that has emerged, there is the potential for tremendous innovations in healthcare. The whole system can be overhauled. Here just a few areas that come to my mind:
1. Medical diagnostics.
2. Rx and drug subscription
3. Smart patient monitoring and alerts
4. Precision medicine
5. Universal access to high quality medicine
6. Automatic surgeries performed by robots

What are the main risks associated with applying deep learning methods into healthcare and medicine?
Sobia: Automating elements of medical practice means physicians will increasingly move away from traditional points of face-to-face patient-physician interaction. This should allow for improvement in quality of time spent with the patient, including more time spent on interpretation, communication and clinical-decision making. Automation of medical diagnosis needs to be accurate and avoid reporting incidental findings that are not backed by proven research. Otherwise the speed and convenience offered by medical AI risks negatively impacting lifestyle, health and reproductive choices if the information and recommendations are misinterpreted in the absence of human input to put these findings into context. Accuracy and reliability is also vital. It is important that medical AI technologies are tested rigorously to prove their purported benefits. Accuracy and reliability go hand in hand with ensuring data security and protecting consumers personal information.

Diogo: A huge problem that medicine will face is people trying to cash in on the recent AI hype, building up unrealistic expectations and burning the trust of both decision makers in the healthcare industry as well as patients. We are on the brink of an AI revolution and we as an industry need to act responsibly, because the actions today could lead to an environment with regulations that prevent people from getting the best care possible.

Vinay: Deep learning has been proven to be working at many situations but we hardly know what is happening inside of these Learning layers. The primary unknown is - we don't know what is happening inside these Neural Nets. We know the outcomes of these networks on a closed test. On a open test, these unknown conclusions could create multiple other conclusions which we might not have anticipated. However, these risks can be mitigated by creating a common test that can filter the odds just like how humans are being tested. But theoretically, any human specialist came from education system would have never answered all the questions correctly. It is only relative - How good are you in your class compared to other students. It does not mean that s/he answered every questions from the professor or posed in a question paper correctly. So, the ‘risk’ is kind of no different to trusting a human expert. The best human experts available are by no means perfect.

Michael: Recommendations need to be precise and accurate. US medicine has been slow to digitise, for cultural and regulatory and cost issues.

Olexandr: My biggest concerns are not connected with AI. I see no danger in wide adoption of deep learning or any other data driven technologies. However, the biggest risks are associated with the privacy of the patients and their data. As we see wider adoption of DL in real life. More people than ever will have access to electronic health records (EHR), electronic medical records (EMR). Hospitals and insurance companies will fully digitize their paper archives. We, as a field, must strive to protect this data, protect privacy of patients and maintain their trust. Other risks are ethical and legal. The same like with self-driving cars, current legal systems in UK and US are not yet adopted for AI. What will happened if patient was misdiagnosed by the AI. Whom to blame? What to do? All of these questions must be resolved.

These experts will be speaking at the RE•WORK Deep Learning in Healthcare Summit in London on 7-8 April 2016. Other speakers include Brendan Frey, President & CEO, Deep Genomics; Daniel McDuff, Principal Research Scientist, Affectiva; Ekaterina Volkova-Volkmar, Researcher, Bupa; Alex Zhavoronkov, CEO, Insilico Medicine; Cosima Gretton, Doctor, Guy's and St Thomas' NHS Foundation Trust; Alejandro Jaimes, CTO & Chief Scientist, AiCure, and more.

Tickets for this event are now limited, for more information and to register please visit the event page here.

Big Data Machine Learning Deep Learning Algorithms Personalised Medicine Deep Learning AI Assistants Healthcare MedTech Genomics Deep Learning in Healthcare Summit Roundtable Q&A


    As Featured In


    Partners & Attendees

    Ibm watson health 3.001
    Rbc research.001
    Mit tech review.001
    Kd nuggets.001
    Maluuba 2017.001
    This website uses cookies to ensure you get the best experience. Learn more