Advanced Analytics Core

Progress

Accomplishments during this reporting period
  • 1
    Engaged with clinical and non-clinical investigators to better understand the capabilities of each technology and the measures of Parkinson disease they capture
  • 2
    Researched and tested multimodal analysis techniques
  • 3
    Published a manuscript on deep phenotyping in Journal of Parkinson Disease

#1

Engaged with clinical and non-clinical investigators

The Advanced Analytics Core met on a monthly basis to further explore each of the technologies (mPower smartphone app, Emerald radio wave device, mc10 wearable sensors, PARK video analytics) used in the ongoing Udall studies and data relationships across platforms. Project team leaders have presented their individual technology, including what data are collected, what the data mean, and how the data is processed. In addition, clinical investigators (PI, Dr. Ray Dorsey and early career investigator, Dr. Chris Tarolli) have joined Core meetings to provide neurology expertise and share perspectives on how the data are meaningful to clinicians. A matrix showing a summary of measures captured by different devices was developed in the planning for the initial SUPER-PD data analyses. Since different devices capture different measures, the matrix will help the Advanced Analytics Core identify which measures are common to one or more of the technologies (e.g. gait, tremor).  A new CHeT Data Scientist, Alex Page, PhD, joined the team to guide efficient data uploads and work with the Core to conduct analyses with the multi-model data. Advanced Analytics Core engineers are actively developing algorithms for extracting and analyzing clinically meaningful data from each technology.

#2

Researched and tested multimodal analysis techniques

In preparation for conducting comparative analyses of the primary data analysis that each of the research project teams will be conducting, Dr. Jiebo Luo, co-leader of the Advanced Analytics Core, and his trainees have been researching and testing multimodal analysis techniques. One initiative has been developing initial binary classifications to reproduce results from a previous mPower study. In addition, the team has been exploring multi-modal learning by late-fusion, hand/pose landmark detection, few-shot learning for disease diagnosis, and multimodal irregular time series analysis techniques.

#3

Published a manuscript on deep phenotyping in Journal of Parkinson Disease

The Journal of Parkinson Disease has accepted our paper, Deep Phenotyping of Parkinson’s Disease for publication. While advances in genetics, imaging, and molecular biology have improved our understanding of the underlying biology of Parkinson’s disease (PD), clinical phenotyping of PD still relies primarily on history and physical examination. These subjective, episodic, categorical assessments are valuable for diagnosis and care but have created gaps in our understanding of the PD phenotype.

In this paper, we explore the concept of deep phenotyping —the comprehensive assessment of a condition using multiple clinical, biological, genetic, imaging, and sensor-based tools —for PD. Sensors can provide objective, continuous, real-world data about the PD clinical phenotype, increase our knowledge of its pathology, enhance evaluation of therapies, and ultimately, improve patient care. We discuss the rationale for, outline current approaches to, identify benefits and limitations of, and consider future directions for deep clinical phenotyping.


The Morris K. Udall Center of Excellence for Parkinson's
Disease Research at the University of Rochester

Udall acknowledgement for publications: “Research reported in this publication was supported by the National Institute Of Neurological Disorders And Stroke of the National Institutes of Health under Award Number P50NS108676. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.”