On Tuesday, May 28th, BCLA hosted its annual healthcare focused event, a panel discussion at UCLA’s Anderson school of management. With both early stage startups and healthcare industry giants GE contributing to the discussion, the event proved to be both informative and exciting. The audience learned about how AI is currently employed within various areas of healthcare and future perspectives of the technology.
The evening began with an introduction from BCLA’s president, Evgeny Bondarenko who highlighted the importance of our organisation's mission as well as upcoming events. Following that, Evgeny introduced the keynote speaker, Mania Samiotaki, product manager for AI at GE Healthcare.
Mania, happy with the fantastic turnout, started her keynote with an in-depth overview of the meaning of AI and the foundations of the technology, dating back to the 1950’s! Mania spoke of the main pillars of AI, including machine learning and deep learning, as well as why the technology is now booming. ‘AI is in our everyday lives!’ Mania told the audience, as she then described the huge advances in GPU power, cloud storage and the plethora of data currently available for AI training. Having set the scene on AI, Mania then turned to the impact the technology is currently having in the field of healthcare. ‘In Europe and North America, we are lucky, there is 1 radiologist for 10,000 patients, in Kenya, this number is 1 for 24,000 patients.’ Mania used this example to highlight the potential for diagnostic errors, estimated between 10-15% which can be drastically reduced or eliminated by using AI. Amazingly, Mania then presented a study by Google to demonstrate that their AI pipelines were more successful at diagnosing cancer than doctors alone, although was quick to mention that AI is not there to replace people in the workplace, rather offer faster, more reliable solutions to workflows. The current investment of $4.3B in >100 start-ups since 2013 with more investment in the pipeline for companies, including GE Healthcare, speaks for itself. Mania then addressed the importance of ethics in AI. She mentioned different challenges that AI faces and the importance of data privacy, although admitted that these aspects were company dependent. ‘At GE we have a checklist. Transparency, no bias and keeping data safe’. Mania concluded her talk by introducing the audience to GE’s Edison platform to specifically aid the fields of X-ray and MRI diagnostics as well as fetal heart ultrasound diagnostics.
After Mania’s keynote, Niyati Thosani from medgenome, our event sponsor, showcased the plethora of services that are offered by their company. Pertinently, Niyati spoke of how AI research is helping mature the fields of genomics and immuno-oncogenics, with the hope of reducing cost and time to make therapies a reality.
Following the these great talks, the AI panel was up next! Panelists, Mania Samiotaki, Adam Simon from Qulabs, Matt Lumpkin from Tidepool and Sumeet Bhatia from Aiva Health answered questions from the panel’s moderator, Amit Rushi from Graftworx, and BCLA advisory board member. Starting with a brief introduction of each panelists background, Amit quickly dove into the meat of specifically who these companies were trying to help, which ranged from nurses, practitioners, patients and caretakers. Matt Lumpkin spoke of how patients need to feel ‘in control’ of their treatments, and how AI can be used to deliver information which then ‘needs an informed decision from the patient’. An important question asked by Amit was how AI was being received by customers and patients and the steps that need to be taken to generate credibility around the technology. ‘Present a great prototype, first impressions are everything’, Adam Simon told the audience. ‘The ultimate goal is to establish autonomous workflows’. Sumeet then explained that AI should be as ‘least disruptive as possible to existing workflows’ and was quick to explain the importance of learning from field personnel, including nurses and patient caretakers. Another question from Amit was where the datasets that AI was being trained on, came from. ‘Data is everywhere’, replied Mania, who described the vast amount of data collected on a vast number of patients in the USA alone. This prompted an important question from an attendee, who asked whether more data was better. This was in turn answered by Adam, who’s company has found that, no, this is not always the case. ‘In some circumstances, the quality of data is more important than the quantity’, framing his answer over unfocused datasets and their resulting skew in results. Potentially, one of the most important questions of the night again came from an attendee who asked the panel about the security of such private data. Sumeet answered by assuring the audience that a large amount of data is actually collected on controlled devices that do not have access to, for example, social media applications, reducing the risk of security breaches. The rest of the panel then discussed how important it was to build secure and trusted platforms for data, nicely tying back into earlier discussion points of how to ensure people embrace AI and trust it’s input into healthcare.
Following this fascinating panel, the rest of the evening played host to a networking mixer which allowed the attendees to continue the great conversations over appetizers and drinks. We would like to thank our fantastic speaker, Mania Samiotaki, our sponsors Medgenome, and our panelists and moderator, Adam Simon, Matt Lumpkin, Sumeet Bhatia and Amit Rushi for their time and dedication to the improvement of healthcare through AI.
If this event or any other event hosted by BCLA has helped you in your career or company, we would love to know! Send us an email with your story to team@bc-la.org and we would love to feature you on our website.