Photo Description: The 2020 Life Sciences Day took place on April 17th and was on the topic of artificial intelligence in healthcare.
By Gisela Cairo, Genome Cell & Developmental Biology PhD student, Soni Lacefield Laboratory and Cullie Poseria, Masters in Business, Marketing
This year’s Life Sciences Day, organized by the Indiana University Kelley School of Business Center for the Business of Life Sciences (CBLS), gave students an opportunity to hear from experts on an industry hot topic, Artificial Intelligence: Changing the Healthcare Landscape. Four panelists came together on Friday, April 17th from various companies to discuss the latest trends during a two-hour webinar that had a great turnout, 61 out of 85 total registrants! Below we’ve included a short recap and link to our question and answer portion.
Photo Description: Participants interact during the school’s annual Life Sciences Day.
The annual event is typically hosted in person, but due to the coronavirus epidemic closures, circumstances led the team to convert the event to an online session. This allowed students from both the undergraduate and graduate business schools to participate, as well as students from the online and science programs. The agenda began with introductions, a group photo and 15-minute presentations from each speaker. The event concluded with a question and answer session paired with a brief recognition for the 29 students that were able to complete their Life Sciences Certificate requirements this year.
Thanks to the event sponsors, the Voris Family Foundation and Workman Family Charitable Fund, the event was free for everyone and went off without a hitch. Speakers Andrew Brockett, the senior director of DMI’s digital tech office for financial services, healthcare and life sciences, Blair Hu, a data scientist from Strata Decision Technology, James Guszcza, Deloitte’s US chief data scientist and Louisa Roberts, the head of commercial strategy at WuXi NextCODE, provided a glimpse into the innovative work being done at their firms.
Photo Description: The diverse speaker lineup brought valuable insight and applications of artificial intelligence to students through presentations and a question and answer session.
Andy Brockett from Digital Management, LLC (DMI) uses artificial intelligence software to help healthcare companies improve access to care for patients. They’ve built a digital transformation consulting business by harnessing a structured, human-centric approach to innovation and scalable optimization. DMI guides companies in determining areas for improvement and in finding appropriate next-generation innovations to address challenges. During his talk, Brockett shared a DMI case study, where his team helped an emergency department at a leading children’s hospital reduce complications with pediatric pneumonia by identifying patterns in patient feedback data, which they translated to solutions of increasing access to care with longer clinic hours and the use of telemedicine but also introducing chatbots to help educate and answer community questions to reduce onset of this preventable condition. This program helped the hospital reduce emergency room visits by nearly 50% and decrease costs by $3 million.
Photo Description: DMI’s Andrew Brockett shared various examples of artificial intelligence in practice to help students understand its potential capabilities.
Blair Hu from Strata Decision Technology shared how the company is applying artificial intelligence to financial data in healthcare. Research shows that only a small number of patients, those with comorbidities, account for most of the cost and strain on our healthcare delivery system. Strata Decision Technology supports healthcare providers with a platform for financial planning, cost accounting, and process improvement to help them provide quality care that is also cost-effective and efficient. They leverage data from across their network of more than 200 clients to benchmark care across providers, tracking financial performance, labor productivity, and cost. They are using both natural language processing and machine learning to apply large-scale data normalization to financial and clinical data across providers. These artificial intelligence tools are used to inform best practices for financial planning and decision support and to identify opportunities for continuous improvement.
Photo Description: At Strata Decision Technology, Blair Hu mentioned that they’ve proposed a new triple aim, not just for healthcare, but for finance overall.
James Guszcza from Deloitte has been able to conduct behavioral hot spotting with ‘big data’ collected on shopping, hobbies, lifestyles and more to fast track people for things such as life insurance underwriting. Hot spotting was also being used for tracking and predicting the spread of COVID-19. Predictive models combined with social determinants of health data helped identify potential high-risk patients in various locations. Guszcza stressed that while algorithms can point to solutions, using the technology was not the hard part. The hardest part was finding ways to create behavioral change during implementation—characterized by a change in habit, judgement and/or decision-making for the people involved. Deloitte combines behavioral change techniques tailored to different types of people with artificial intelligence algorithms, to target and use a process called precision nudging. Jim emphasizes use of ‘behavioral nudge’ or ‘choice architecture’ that changes small things in the environment (the ‘nudge’) to help people alter their behavior, for example to follow healthier habits. Making it purposeful, easy and convenient for people to change is critical to success.
Photo Description: Deloitte’s James Guszcza shares slides from a partnership with Penn Medicine, where they experimented and determined the most effective ‘nudges’ for different types of people.
WuXi NextCODE (WXNC) uses genomic data and AI to aid biopharma and academic partners with genomic insights. They develop disease-specific data sets to better understand disease biology in order to identify potential therapeutic targets and or biomarkers. Louisa Roberts spoke about how she liaises between the firm’s clients and the WXNC data science team, led by Tom Chittenden. Roberts says that “AI is a tool that can solve infinite problems, but the ‘trick’ is to bring the right tools to the right problems.” Based on literature, she mentioned how a drug target is two times more likely to achieve FDA-approval if it has genetic validation. WXNC believes that using whole genome sequencing with deep phenotypic data will also help to reduce overall costs by speeding drug development to market. The team uses deep machine learning to identify patterns in multi-omic data sets which are then put in context with existing clinical literature, biology and domain expertise. She finished her talk by displaying a beautiful Mass Cytometry image. Artificial intelligence was used to identify a disease-prone cluster of cells as they transformed over time, the imaging mass cytometry showed visually the biological signal that the AI revealed in the data. She emphasized how mathematics and statistics are very important in the process of data analysis since they help point to what is biologically significant.
Photo Description: Louisa Roberts shares image analysis of the disease driving cells identified by the WuXi NextCODE artificial intelligence.
During the question and answer session, there were some interesting highlights as well.
Q: “What are some best practices for business and data science teams to work together?
Blair answered that it’s important to simply communicate the value of the data to the business side. Receptivity comes from reducing the complexity behind the analysis.
Louisa also thought that educating, translating and advocating to stakeholders are essential and pointed out how simplifying complex processes to something clients can relate to and apply is extremely beneficial.
Jim agreed with Louisa, quoting “data science is a team sport” and emphasizing how important it is to communicate the math and science in a way clients can easily understand and apply.
Q: “What are some of the human components behind the use of artificial intelligence?”
All the speakers agreed that some of the human components that cannot replace AI are reasoning and ethical application, ability to innovate, to analyze data without biases and to critically think, judge and use findings.
Leave a Reply