ICHSM 2024

Tutorial Information

Tutorial Information on 7 September, ICHSM 2024

Time: 13:10-17:25 (break time: 15:10-15:25)

Registration Deadline: August 5, 2024

Tutorial Theme: Process Mining for Healthcare: From Theory to Practice with pMineR

pMineR
R libraries for Process Mining for Healthcare
The pMineR project was born in 2016 and presented the first time at AIME 2016, in Vienna. We are glad for the opportunity to be here now to present the current state of this never-ending development, share visions and perspective and collects all the potential contributes of users and interested colleagues.
pMineR is an R library specifically designed to support Process Analysts to work in the clinical domain, providing Process Discovery algorithms, tools for Conformance Checking, Trace and Event Log analysis and methods for representing and working with given institutional processes, Consensus flow, Clinical Protocols, etc.. The library implements those features considering the most frequent statistical tools used in healthcare, such as non-parametric tests (Chi-Square, Fisher's Exact test), Surivival Analysis (Kaplan Meier curves, log-rank tests) and also Machine Learning methods and predictors to estimate the most relevant clinical outcomes in terms of events or times. It also implements an easy to use internal language for the interpretable representation of clinical guidelines and subsequent compliance analysis (crossing the domains of Process Mining in Healthcare and Clinical Computer Interpretable Guidelines)
For more info, visit : https://github.com/PMLiquidLab/pMineR.v046

 

Tutorial Chair: Prof. Roberto Gatta, Department of Clinical and Experimental Sciences, Università degli Studi di Brescia, Italy

 

Department of Information Engineering, Università degli Studi di Padova (IT)

Department of Electrical, Computer and Biomedical Engineering at the University of Pavia (IT)

Department of Industrial and Information Engineering, Università degli Studi di Pavia (IT)

School of Computing and Engineering, University of Huddersfield (UK)

Department of Oncology, Lausanne University Hospital, Lausanne (CH)

Tutorial Organizers: Erica Tavazzi, Department of Information Engineering, Università degli Studi di Padova (IT), Italy; Lucia Sacchi, Department of Electrical, Computer and Biomedical Engineering at the university of Pavia, Italy; Arianna Dagliati, Department of Electrical, Computer and Biomedical Engineering at the university of Pavia , Italy; Mauro Vallati, School of Computing and Engineering, University of Huddersfield, UK; Michel Cuendet, Department of Oncology, Lausanne University Hospital, Lausanne, Swiss

About the lecturers: Roberto worked at the Gemelli Hospital in Rome, at the Centre Hospitalier Universitaire Vaudoise (CHUV, Lausanne) and currently at the Department of Clinical and Experimental Sciences, Università degli Studi di Brescia. He works in the field of Process Mining for Healthcare since 2016 and is the most active developer and current maintainer of the package pMineR.

This Tutorial is designed to provide a general overwiew on pMineR, and introduce the participants to the main modules, by short hands-on sessions, to cope with Process Discovery, Conformance Checking and Statistical Analysis of paths, clinical guidelines, events and timing. The Tutorial will also exploit invited speakers to present their experience on real-world data analysis and open projects based on pMineR. Finally, an open round-table will be the opportunity to share opinions, ideas about the topic or a further evolution of the libraries. An indicative program is shown below :

Frontal Lesson Modules ( 1h 30', Roberto Gatta )：

General introduction
• Event Log data Loading, internal structure, events, time ( pMineR::dataLoader() )
• Performing queries and Quality-of-data oriented trace inspection ( pMineR::QOD() )
• Process Discovery with First Order Markov Models and CareFlow Miner
• Clinical Guideline Conformance Checking with Pseudo Workflow language ( pMineR::confCheck_easy() )
• Short lectures kept by invited external speakers (1 hour )
• Round Table ( 30' ):
• Freewheeling discussion about the current issues or future opportunities of the discipline and/or the project. For esample: empowering the PD with more ML? PD-oriented scheduling and planning DSS, are they possible? Engaging MDs: winning and losing strategies? PM4HC and Semantic web technologies: which possible marriage? Big Data = Big Problems: can we scale PM analysis? Federated PM to cope with Patient's privacy and data ownership, in multicentric clinical studies? .... and any other Wild and Crazy ideas!