ICDM-TMDM 2019 Workshop 11/8-11 in Beijing

Beijing China

ICDM-TMDM 2019 Workshop 11/8-11 in Beijing

Translational Multimedia Data Mining for AI-Based Medical Diagnostics—Bridging Digital Intelligence with Clinical Practices (TMDM for short) is a WORKSHOP collocated with ICDM 2019 (November 8 – 11, 2019) in Beijing, China.



Recent developments of health informatics and digital medical diagnostics have accumulated a large amount of multimedia data. Typical examples include the face/gesture videos, medical scan image, sensor signals, and multimodal medical databases. The volume and complexity of these data provide significant data mining challenges as well as the opportunity to develop robust digital tools with clinical potentials. This workshop aims at bringing together researchers from medical data analysis, medical signal processing, health informatics, clinical research, statistical pattern recognition, machine learning, and artificial intelligence to share their recent progress and synergies. Emphasis will be on the potential towards translational medical research, clinical research and practice, AI-based diagnostic tools and databases, and a clinical-inclined integration of the data mining, bioinformatics, medical science, and clinical research areas.


Topics of Interests

The topics of interests for the workshop include, but not limited to the following:

  • Deep learning based multimedia analysis tools and applications for healthcare
  • Deep learning based data mining tools for multimedia data retrieval
  • Physiological and psychological sensing tools for healthcare data capturing, pre-processing, storage, retrieval, annotation, and utilization
  • Clinical protocols and practices for multimedia content modeling and interpretation in clinical context
  • AI-based diagnostic tools for improving medical operation efficiency and effectiveness
  • Legal, ethical, and policy aspects of health informatics and AI medical data mining tools
  • Operational and clinical trials of health informatics and machine-aided diagnostic tools
  • Human factors, user-centered designs, integrative visualizations of data mining applications on medical science


Workshop Agenda

Keynote Speakers

Daniel MessinigerKeynote 1:   Friday, November 8  | 8:10am – 9:05am | Room 301AB
Data Drive Development—Objective Phenotyping of Autism Spectrum Disorder

(Daniel Messinger, University of Miami)

ABSTRACT  How can we harness computational approaches to early social interaction create new insights into typical development and communication disorders such as autism? Employing a dynamic systems framework to developmental process provides structure to new initiatives in the interface between data mining and developmental science. Unsupervised learning highlights the role of positive emotion in early interaction. Signal processing of video and vocal interaction is shedding light on the diagnosis of autism. Sequential analysis of patterns of head movement suggest autism diagnosis differentiation. Radio frequency identification of children’s classroom movement and automated detection of vocal interaction yield classroom networks and suggest how language development occurs. Despite these advances, computational approaches are challenged to provide insight into behavioral phenomena. The talk will consider the strengths, challenges, and future of computational approaches to understanding the autism phenotype.

SPEAKER BIO  Daniel S. Messinger, PhD. Dr. Messinger is an interdisciplinary developmental psychologist, and the author of over 120 scientific publications appearing in journals such as PLoS ONE, Developmental Science, and Molecular Autism. He has experience leading longitudinal research initiatives funded by the National Institutes of Health, the National Science Foundation, and the Institute of Education Sciences. Dr. Messinger investigates the temporal dynamics of communication to understand how infants and children’s social, emotional, and language development. He uses machine learning to paint an objective picture of children’s interaction and employs computational models to make sense of the resulting big behavioral data. Dr. Messinger works with children affected by autism spectrum disorder (ASD), hearing loss, and poverty. By understanding interaction, he seeks to fosters pathways to healthy development. Specific projects include the emergence of secure attachment, objective measures of autistic behavior, and language and proximity/orientation networks in inclusive classrooms.


Masaharu Goro

Keynote 2:   Friday, November 8  | 9:05am – 10:00am | Room 301AB
Scaling-Up Heterogeneous Waveform Clustering for Long-Duration Waveform Recording and Analysis

(Masaharu Goto, Keysight Technologies International Japan)

ABSTRACT  High speed ADC (Analog to Digital Converter) and huge data storage are commodities in today’s electronic products. The volumes of data those devices can produce are extremely large. For example, recording at 10MSaPS (Mega Sample Per Second) for 24 hours produces 1TB of data. Combining the real-time processing requirements, the related application scenarios pose many “big data” challenges. In this talk, I will present a novel heterogeneous waveform clustering framework for exploring and analyzing such data, which allowing effective user exploring and data organization. This framework implements real-time tagging to enable low-cost instrument measurement platform to handle big data waveform recordings. The real-time tagging algorithm creates meta-data for fast access and interactive clustering of waveforms at large analysis scale while continuously recording high-speed sampling data for hours. This framework provides efficient waveform data capturing, organization, and analytics, and user interaction solutions to many modern big-data application, e.g., data segments with specific patterns in a huge waveform database can be quickly retrieved without large computer workstation clusters or high-performance computing resources. I will also cover several application examples that tackle the challenges of the recent large-scale deployment of IoT (Internet of Things) devices and their related data science innovations.

SPEAKER BIO Masaharu Goto is a Principal Research Engineer in Keysight Technologies’ Electronic Industry Solution Group. He co-developed the ROOT/CINT scientific data analysis framework (https://root.cern.ch/; https://root.cern.ch/cint) for CERN (European Organization for Nuclear Research https://home.cern/), which was the world’s first big data processing framework for the world’s largest particle physics experiments’ data recording and analysis. He provided C++ interpreter for seamlessly connecting interactive big data exploration and high-performance computing for enabling the analysis on CERN’s Large Hadron Collider (LHC: the world’s largest particle accelerator https://home.cern/science/accelerators/large-hadron-collider). At Keysight, he spearheaded the research and development of various test and measurement systems for big measurement data environments. These systems enable the massive parametric measurements for the most advanced semiconductor research and high-volume production. His current research projects combines big data analytics with real-time data analysis frameworks to implement interactive and productive instrument platforms for signal processing and data mining practitioner on long-duration plus high-sampling rate scenarios. His research interests include instrument measurement, big data, data mining, and exploring the broader frontiers of computer science.



Paper 1:   Friday, November 8  | 10:30am – 10:45am | Room 301AB
Implementation of Mobile-Based Real-time Heart Rate Variability Detection for Personalized Healthcare
Luis Quintero (Stockholm University); Panagiotis Papapetrou (Stockholm University); John Edison Munoz Cardona (University of Waterloo); and Uno Fors (Stockholm University)

Paper 2:   Friday, November 8  | 10:45am – 11:00am | Room 301AB
Multi-Scale Sequential Pattern Discovery and Alignment for Long-Duration Waveform Similarity Quantification and Interpretation
Masaharu Goto (Keysight Technologies International Japan); Naoki Kobayashi (Keysight Technologies International Japan); Gang Ren (University of Miami); and Mitsunori Ogihara (University of Miami)

Paper 3:  Friday, November 8  | 11:00am – 11:15am | Room 301AB
Survival of the Fastest: Using Sequential Pattern Analysis to Measure Efficiency of Complex Organizational Processes
Joseph Johnson (University of Miami); Raju Parakkal (Philadelphia University); Sherry Bartz (University of Miami); Gang Ren (University of Miami); and Mitsunori Ogihara (University of Miami)

Paper 4:   Friday, November 8  | 11:15am – 11:30am | Room 301AB
Deep Learning-Based Multimedia Data Mining for Autism Spectrum Disorder (ASD) Diagnosis
Saad Sadiq (University of Miami- Miller school of medicine); Michael Castellanos (GreenPath Financial ); Jacquelyn Moffitt (University of Miami ); Mei-Ling Shyu (University of Miami); Lynn Perry (University of Miami); Daniel Messinger (University of Miami)

Paper 5:   Friday, November 8  | 11:30am – 11:45am | Room 301AB
Audio-Based Group Detection for Classroom Dynamics Analysis
Yudong Tao (University of Miami, USA); Samantha Mitsven (University of Miami); Lynn Perry (University of Miami); Daniel Messinger (University of Miami); Mei-Ling Shyu (University of Miami)



Justine Cassell, School of Computer Science, Carnegie Mellon University, USA

Jeffrey Cohn, Department of Psychology, University of Pittsburgh, USA

Zakia Hammal, Robotics Institute, Carnegie Mellon University, USA

Katherine Martin, Tobii Pro, USA

Daniel Messinger, Department of Psychology, University of Miami, USA

Mitsunori Ogihara, Department of Computer Science, University of Miami, USA

Gang Ren, Institute for Data Science and Computing, University of Miami, USA


Program Committee

Oya Aran, IDIAP Research Institute, Switzerland

Mohamed Chetouani, UPMC, France

Jeffrey F. Cohn, University of Pittsburgh, USA

Zakia Hammal, Robotics Institute, Carnegie Mellon University, USA

Hongying Meng, Brunel University London, UK

Mitsunori Ogihara, Department of Computer Science, University of Miami, USA

Gang Ren, Institute for Data Science and Computing, University of Miami, USA

Akane Sano, Rice University, USA

Babak Taati, University Health Network, Canada

Lijun Yin, Department of Computing Science, SUNY at Binghampton University, USA


Related Links

The IEEE ICDM 2019 Conference main page: http://icdm2019.bigke.org/

The Submission website: https://cmt3.research.microsoft.com/TMDM2019/Submission/Index




Mitsunori Ogihara



Mitsunori Ogihara | m.ogihara@miami.edu




Gang Ren



Gang Ren | gxr467@miami.edu