InnovationWell: Exploring AI and Machine Learning approaches to Precision Medicine and Healthcare
The unprecedented quantities of data now being generated by medical research and healthcare are far beyond human management. To turn this torrent of Big Data into Smart Data, scientists are increasingly using Artificial Intelligence and Machine Learning.
But there have been growing pains. Recent claims, for example, that IBM’s supercomputer ‘Watson for Oncology’ recommended “unsafe and incorrect” cancer treatments are a reminder that applying AI and ML to medical research and healthcare depends both on the datasets selected and training in interpretation and use. Outputs are only as good as inputs.
At two InnovationWell events in October, several preeminent speakers will look at the issues around AI and ML from their own particular perspectives. In 15-minute talks, the speakers will explain the essential context of their work, as well as their own particular conundrums. They will, then, join participants in small groups to discuss and explore. What kind of ideas and innovation can come out of such an event? Be a part of InnovationWell and see.
On October 18 at Philadelphia University Place 2.0:
Barry Hardy, CEO of Douglas Connect, will open the panel with a preamble on why now is the time to create new Social, Technical, Scientific, Legal and Regulatory Structures, as well as moderating questions and answers.
Vertex Labs’ Lucas Beauchemin, Vice President of Research and ML Engineer, Dawer Jamshed, Founder, and Tom Bell, CEO, will explain how their company is applying Machine Learning algorithms to smaller data sets, in particular how Vertex’s approach is achieving dramatic improvements over other methods used in the field of chemometrics relating to drug measurements. “One subset of Machine Learning, Deep Learning, further facilitates the ability to elucidate insight without heavy feature engineering from experts,” says Tom Bell.
Michael N. Liebman, Managing Director and Co-Founder of IPQ Analytics, will talk about the real-world complexity of building models of disease based on his own collaboration with medical doctors. “Precision medicine really means imprecise medicine,” he says, “because the vast majority of people who talk about PM don’t understand how real medicine is practiced.”
Jaclyn N. Taroni, Data Scientist at the Childhood Cancer Data Lab (CCDL), will talk about how she is applying Machine Learning to sub-biology to search for a cure for childhood cancer. “We want to give researchers the power to use the vast amounts of publicly-available data in a meaningful way,” she says. “Machine learning can take an off-the-shelf-gene, extract the data, do machine learning on it and then transfer that learning to what has already been learned about a data set and apply it.”
And on October 22 at Cambridge Innovation Center:
Barry Hardy, CEO of Douglas Connect, will open the panel and moderate questions and answers.
Bruce Kristal, Associate Professor, Brigham and Women's Hospital
Tonya Hongsermeier, VP and Chief Medical Information Officer at Lahey Health
Laszlo Urban, Global Head of Preclinical Safety Profiling, Novartis Institutes for Biomedical Research
The unprecedented quantities of data now being generated by medical research and healthcare are far beyond human management. To turn this torrent of Big Data into Smart Data, scientists are increasingly using Artificial Intelligence and Machine Learning.
But there have been growing pains. Recent claims, for example, that IBM’s supercomputer ‘Watson for Oncology’ recommended “unsafe and incorrect” cancer treatments are a reminder that applying AI and ML to medical research and healthcare depends both on the datasets selected and training in interpretation and use. Outputs are only as good as inputs.
At two InnovationWell events in October, several preeminent speakers will look at the issues around AI and ML from their own particular perspectives. In 15-minute talks, the speakers will explain the essential context of their work, as well as their own particular conundrums. They will, then, join participants in small groups to discuss and explore. What kind of ideas and innovation can come out of such an event? Be a part of InnovationWell and see.
On October 18 at Philadelphia University Place 2.0:
Barry Hardy, CEO of Douglas Connect, will open the panel with a preamble on why now is the time to create new Social, Technical, Scientific, Legal and Regulatory Structures, as well as moderating questions and answers.
Vertex Labs’ Lucas Beauchemin, Vice President of Research and ML Engineer, Dawer Jamshed, Founder, and Tom Bell, CEO, will explain how their company is applying Machine Learning algorithms to smaller data sets, in particular how Vertex’s approach is achieving dramatic improvements over other methods used in the field of chemometrics relating to drug measurements. “One subset of Machine Learning, Deep Learning, further facilitates the ability to elucidate insight without heavy feature engineering from experts,” says Tom Bell.
Michael N. Liebman, Managing Director and Co-Founder of IPQ Analytics, will talk about the real-world complexity of building models of disease based on his own collaboration with medical doctors. “Precision medicine really means imprecise medicine,” he says, “because the vast majority of people who talk about PM don’t understand how real medicine is practiced.”
Jaclyn N. Taroni, Data Scientist at the Childhood Cancer Data Lab (CCDL), will talk about how she is applying Machine Learning to sub-biology to search for a cure for childhood cancer. “We want to give researchers the power to use the vast amounts of publicly-available data in a meaningful way,” she says. “Machine learning can take an off-the-shelf-gene, extract the data, do machine learning on it and then transfer that learning to what has already been learned about a data set and apply it.”
And on October 22 at Cambridge Innovation Center: