“Machine learning is changing everything — except maybe healthcare,” said MIT professor, John Guttag at the Big Data and Healthcare Analytics Forum on Oct. 24, 2016.
While Artificial Intelligence or machine learning is already transforming many industries, healthcare providers have done much less with these technologies. I have written up to a hundred stories covering these technologies and the impact I’ve observed so far is enough to prove the hypothesis that machine learning can truly provide convenience to human lives. How? This story will answer that.
Machine learning in healthcare has started changing a lot of things and the revolution we’ve promised through this technology is here. During my recent trip to S.F Bay Area, I’ve had a chance to meet an entrepreneur who is already challenging the statement of Prof. John Guttag.
Richard Lin is an entrepreneur in the Bay Area who is working on a healthcare startup, Thryve. It provides a gut testing kit paired with personalized probiotics to provide knowledge about the microbes inside a human body. His venture aims to improve the health of people by utilizing machine learning and Natural Language Processing (NLP) on 35
00+ microbiome studies to provide tips, recommendations, and actions for improvement. What Richard plans to do ahead is in the following interview. Have a read!
Hira: What does Thryve do? How would you explain the idea behind it?
Richard: Thryve helps people learn about the microbes inside their body in order to improve health. We offer a gut wellness test kit and based on the results deliver high-quality pro/prebiotics to their door. Results are sent to customers with information on the various health impacts of the bacteria levels found in their stomach and diet and lifestyle recommendations. Each report also features an overall “Gut Wellness Score” which allows subscribers to track their progress over time.
Hira: What made you choose probiotics for the well-being of humans?
Richard: Using antibiotics allows us to effectively control microbial infections by killing them. However, we have actually been co-evolving with microbes and with recent advancements in biology and DNA sequencing new discoveries are made about their essential role on our health. Probiotics have been known to be beneficial for our health, but we believe what we know is just the tip of the iceberg, and there is so much more potential for them to help people. Although the science is still in its infancy we’ve seen probiotics exert positive effects on mood, muscle growth, inflammation, lincreasing pain thresholds, prenatal growth, and many others.
Hira: Do you only test for microbiome or do you provide nutritional recommendations too?
Richard: Currently, we test the bacterial microbiome in the human body. We also provide dietary and lifestyle recommendations via our reports.
Hira: Coming to the main point, how are you incorporating Machine Learning into it?
Richard: We use machine learning, deep learning, and natural language processing to summarize contextual information from 3500+ microbiome studies and our customer’s profile data. We then use our dashboards and chatbots to deliver relevant and personalized wellness recommendations to our customers in a user-friendly way. Think of it as your very own personal robot scientist and nutritionist working in conjunction to improve your wellness.
Hira: How far has Machine Learning advanced over recent years and what impact it had on healthcare industry?
Richard: We’ve seen some huge leaps in machine learning across consumer Internet companies. For example, Google photos with deep learning to auto-tag and categorize your photos and Facebook’s computer vision for facial recognition. However, we’ve yet to see a huge leap forward with the healthcare industry. Intrinsically, healthcare is a very difficult problem to tackle considering the stakes are much higher. For instance, tagging a photo incorrectly at scale isn’t quite that large of a problem. However, if errors happen with prediction models for healthcare, serious issues arise and the repercussions can be detrimental to a person’s well-being.
Hira: What other technologies are you planning to incorporate in the near future?
Richard: Chatbot interface, Bioinformatics, Machine Learning.
Hira: Many companies are incorporating Chatbots into the healthcare system. Are you planning the same?
Richard: Absolutely. Starting from last year (Q4’15) chat has surpassed social networks in terms of engagement with users. Thus, we are seeing a huge trend in how users interact with applications and the internet. Chatbots will be a key component in how Thryve will extract meaningful customer insights in an intuitive and natural way about diseases, symptoms, supplements, diet, sleep, etc. We can remove the barrier of cumbersome questionnaires and surveys and build an interactive way for users to provide their data to help advance science.
Hira: How are you planning to incorporate big data into the business?
Richard: We are in the business of making sense of microbial genomics. Intrinsically, removing the noise and making sense of these bacterial genes requires large amounts of data flowing through our proprietary bioinformatics pipeline and our algorithms match and summarize research studies to provide relevant information to our customers.
Hira: What feedback have you received so far on Thryve?
Richard: We were able to build up initial traction with customers in closed Facebook groups dealing with chronic illnesses. The reception has been tremendous and our early adopters are super excited about our product. That said, those who are already healthy have not been as keen to understand the importance of how the microbiome affects their wellness (better sleep, clearer skin, less fatigue, improved mood) and I believe there is an educational process needed to help everyone understand the relevancy.
Hira: Is Google DeepMind your competitor?
Richard: No, I see them as our partners.
Hira: How are you planning to position Thryve in the coming years?
Richard: Thryve’s mission has always been to solve disease associated with the microbiome. However, it’s naive to think that the microbiome is the only area that affects disease states. That said, we envision our platform will target many areas of wellness across humans, animals, agriculture, food. We hope to bring wellness to all areas of life.
P.S. Any questions you want to ask Richard? Drop them to me at firstname.lastname@example.org and I’ll make sure to have them answered!