Agenda

Date and TimeTitle
Nov 2, 2020 (Eastern)
9:00am - 9:33am
Keynote -Using Healthcare Data to Advance Precision Medicine

Our modern healthcare system is characterized by systems that generate unprecedented quantities of data. Estimates are that the average hospital in the US generates more than 667TB of data annually and that as much as 30% of the entire world’s stored data is generated in the health care industry. Despite this deluge of data, its effective use remains an ongoing challenge and meeting the promise of precision medicine—matching each patient to the best therapy based on the unique characteristics of his or her own disease—is far from being realized. While AI and machine learning have often been proposed as a solution, these methods have also failed to deliver on their promise. However, there is hope. Methods that can deal with the complexity of data and impose biologically derived constraints have the potential to dramatically improve our use of the available data and advance both our understanding of disease and our effective delivery of care.

Nov 2, 2020 (Eastern)
9:40am - 10:05am
Pioneering Precision Medicine in Rheumatoid Arthritis

The presentation will discuss the development and commercialization of the first anti-TNF therapy predictive diagnostic test in RA. Scipher Medicine applied its Network Medicine platform to discover and validate TNFi response biomarkers in a large patient population from the Corrona registry.

Nov 2, 2020 (Eastern)
10:10am - 10:27am
Keynote - Artificial Intelligence and Precision Medicine

In this talk we describe various AI enabled clinical solutions with different levels of complexity in terms data requirement and complexity facilitating various decision making processes. These systems are aimed to expand precision medicine, specifically in four areas of improving diagnostic accuracy, reducing unwarranted variations, personalizing when it matters, and advancing therapy outcomes.  

Nov 2, 2020 (Eastern)
10:35am - 11:18am
Keynote - Machine Learning – Getting Up Close and Personal with Precision Medicine: Implications of Single-Subject Studies (S3) for Therapeutic Classifiers

This presentation will review recent innovative translational computational biology and bioinformatics analytics that substantially advanced our accuracy in predicting effect sizes and statistical significance in single-subject studies (S3) when used jointly with machine learning algorithms. S3 can potentially increase clinical trials’ efficiency through reducing clinical trial cohort sizes by more than 50%. Indeed, S3 studies enable paradigm-shifting analytics. Each subject serves as their own case and control, generating significance metrics (pvalues) associated to the magnitude of their altered pathophysiological mechanisms (effect size). These provide substantially smaller and unbiased list of biomechanisms likely relevant to the patient response than conventional machine learning classifier features (inputs).

We will demonstrate the increase accuracy of these principles in simulations of biomarker discovery and in one published clinical trial validation where ML classifiers of therapeutic response can be derived from a small set of twenty subjects and led to identify children prone to asthma exacerbations in another independent validation.

Nov 2, 2020 (Eastern)
11:30am - 11:57am
ML / AI Trends in Diagnostics for 2020 and Beyond

2020 has been another strong year for machine learning and artificial intelligence within healthcare, and the acceleration of progress in ML / AI is clearly going to continue. This presentation will cover 3 major trends that are expected to shape the space in the next decade.  1) ML / AI will augment, not replace providers within the clinical setting.  2)  Both governmental and medical bodies will increase their involvement in providing both guidance and regulation to the use of ML and AI in healthcare.  3) Big tech will continue to influence and disrupt the space through multiple routes.

 
 
Nov 2, 2020 (Eastern)
12:10pm - 1:00pm
Leveraging AI in Digital Pathology to Improve Cancer Phenotyping

Digital pathology is well poised to make significant impacts to the field of precision medicine. In this talk the presenters will discuss the how digital pathology has been employed at Bristol Myers Squibb to extract information from routine pathological specimens in the context of oncology. Recent advances in methodologies will be summarized in regards to end-to-end deep learning with their potential impact on research efforts into solid tumors.

 

Nov 2, 2020 (Eastern)
1:00pm - 1:19pm
Keynote - Transforming the Medical Research Enterprise with Decentralized Clinical Trials

The practice of clinical trials has evolved very little in the past century. New technologies now hold promise for substantial improvements in data, privacy, security, and pace of learning. We describe the implications of mobile phones (edge computing), federated learning, and other new technologies on the research enterprise.