pith. sign in

arxiv: 1907.07669 · v1 · pith:PLVRJ6AJnew · submitted 2019-07-16 · 📊 stat.AP

Sequential Pattern mining of Longitudinal Adverse Events After Left Ventricular Assist Device Implant

Faezeh Movahedi , Robert L. Kormos , Lisa Lohmueller , Laura Seese , Manreet Kanwar , Srinivas Murali , Yiye Zhang , Rema Padman
show 1 more author
James F. Antaki
This is my paper

Pith reviewed 2026-05-24 20:17 UTC · model grok-4.3

classification 📊 stat.AP
keywords LVADadverse eventssequential pattern mininghierarchical clusteringMarkov modelingINTERMACSheart failuredevice complications
0
0 comments X

The pith

Seven groups of ordered adverse event chains follow LVAD implantation in registry data.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper converts records of 58,575 adverse events from 13,192 LVAD patients into chronological sequences per patient. It then applies hierarchical clustering to group patients with similar sequences and Markov modeling to extract the most probable ordered chains within each group. The result is seven distinct trajectories, each dominated by a recurring type and order of events such as recurrent bleeding or neurological dysfunction leading to death. A reader would care if these ordered patterns reflect real clinical interdependence rather than isolated events, because that would open routes to earlier prediction and targeted prevention after device implant.

Core claim

The mined results indicate the existence of seven groups of sequential chains of AEs, characterized by common types of AEs that occurred in a unique order. The groups were identified as: GRP1: Recurrent bleeding, GRP2: Trajectory of device malfunction & explant, GRP3: Infection, GRP4: Trajectories to transplant, GRP5: Cardiac arrhythmia, GRP6: Trajectory of neurological dysfunction & death, and GRP7: Trajectory of respiratory failure, renal dysfunction & death. These patterns of sequential post-LVAD AEs disclose potential interdependence between AEs and may aid prediction, and prevention, of subsequent AEs in future studies.

What carries the argument

The three-step procedure of converting each patient's adverse events into a single chronological sequence, grouping similar sequences by hierarchical clustering, and extracting temporal chains within each cluster via Markov modeling.

If this is right

  • The seven groups each follow a unique recurring order of adverse events.
  • The chains indicate potential interdependence among events after LVAD implant.
  • The identified patterns could support future efforts to predict later events from earlier ones.
  • The patterns could support future efforts to prevent later events from earlier ones.
  • Each group links to distinct clinical outcomes such as transplant, explant, or death.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Clinicians could monitor for the first event in a known chain to trigger earlier interventions aimed at blocking progression.
  • Risk models for LVAD patients might improve by incorporating event order rather than counting events independently.
  • The same sequencing-plus-clustering approach could be tested on other implanted devices or on non-device heart-failure cohorts to check whether similar ordered chains appear.
  • Validation would require checking whether patients assigned to one group show measurably higher rates of the predicted later events than patients in other groups.

Load-bearing premise

Converting recorded adverse events into patient sequences and then clustering plus Markov modeling will surface real clinical interrelations rather than artifacts of recording practices or algorithm choices.

What would settle it

Re-running the identical three-step procedure on the same INTERMACS data but with a different linkage method in clustering or a different order threshold in the Markov step produces groupings whose chains do not match the seven reported trajectories.

Figures

Figures reproduced from arXiv: 1907.07669 by Faezeh Movahedi, James F. Antaki, Laura Seese, Lisa Lohmueller, Manreet Kanwar, Rema Padman, Robert L. Kormos, Srinivas Murali, Yiye Zhang.

Figure 1
Figure 1. Figure 1: Illustration of HeartMate 3 LVAD from Thoratec [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: General work-flow: 1. Data preprocessing:Forming patients’ sequences. 2. Hierarchical clustering:cluster patients into [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Color code for 15 types of AEs and final outcomes [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: The proportions of various types of AE in all the 13,192 [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 3
Figure 3. Figure 3: in which the proportions of each category of AE was [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 6
Figure 6. Figure 6: length of AE sequences from 1 to 15. (Length of 16-32, [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The dissimilarity scores distribution of all the sequences [PITH_FULL_IMAGE:figures/full_fig_p005_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: First step of step-wise cluster evaluation: evaluation of the two-cluster solution [PITH_FULL_IMAGE:figures/full_fig_p006_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Regression tree showing groups formed at each step [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: MM of GRP1: Recurrent bleeding (n= 862 patients) [PITH_FULL_IMAGE:figures/full_fig_p008_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: MM of GRP2: Trajectory of device malfunction & [PITH_FULL_IMAGE:figures/full_fig_p008_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: MM of GRP3: Infection (n= 3,438 patients) [PITH_FULL_IMAGE:figures/full_fig_p009_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: MM of GRP4: Trajectories to transplant (n= 3,302 [PITH_FULL_IMAGE:figures/full_fig_p009_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: MM of GRP5: Cardiac arrhythmia (n= 1,275 patients) [PITH_FULL_IMAGE:figures/full_fig_p009_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: MM of GRP6: Trajectory of neurological dysfunction [PITH_FULL_IMAGE:figures/full_fig_p010_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: MM of GRP7: Trajectory of respiratory failure & [PITH_FULL_IMAGE:figures/full_fig_p010_16.png] view at source ↗
read the original abstract

Left ventricular assist devices (LVADs) are an increasingly common therapy for patients with advanced heart failure. However, implantation of the LVAD increases the risk of stroke, infection, bleeding, and other serious adverse events (AEs). Most post-LVAD AEs studies have focused on individual AEs in isolation, neglecting the possible interrelation, or causality between AEs. This study is the first to conduct an exploratory analysis to discover common sequential chains of AEs following LVAD implantation that are correlated with important clinical outcomes. This analysis was derived from 58,575 recorded AEs for 13,192 patients in International Registry for Mechanical Circulatory Support (INTERMACS) who received a continuousflow LVAD between 2006 and 2015. The pattern mining procedure involved three main steps: (1) creating a bank of AE sequences by converting the AEs for each patient into a single, chronologically sequenced record, (2) grouping patients with similar AE sequences using hierarchical clustering, and (3) extracting temporal chains of AEs for each group of patients using Markov modeling. The mined results indicate the existence of seven groups of sequential chains of AEs, characterized by common types of AEs that occurred in a unique order. The groups were identified as: GRP1: Recurrent bleeding, GRP2: Trajectory of device malfunction & explant, GRP3: Infection, GRP4: Trajectories to transplant, GRP5: Cardiac arrhythmia, GRP6: Trajectory of neurological dysfunction & death, and GRP7: Trajectory of respiratory failure, renal dysfunction & death. These patterns of sequential post-LVAD AEs disclose potential interdependence between AEs and may aid prediction, and prevention, of subsequent AEs in future studies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The manuscript presents an exploratory analysis of sequential patterns in adverse events (AEs) after LVAD implantation, using 58,575 AEs from 13,192 patients in the INTERMACS registry (2006-2015). It applies a three-step pipeline: (1) converting per-patient AEs into chronological sequences, (2) hierarchical clustering to group patients with similar sequences, and (3) Markov modeling to extract temporal chains within each cluster. The central claim is that this yields seven distinct groups with unique AE orders (recurrent bleeding; device malfunction & explant; infection; trajectories to transplant; cardiac arrhythmia; neurological dysfunction & death; respiratory failure, renal dysfunction & death) that disclose AE interdependencies and may aid prediction.

Significance. If the seven groups prove robust, the work would offer a data-driven view of AE interdependence beyond isolated-event studies, leveraging a large registry sample to generate clinically relevant hypotheses. The approach is novel in this domain, but the lack of reported validation means the claimed patterns cannot yet be distinguished from artifacts of follow-up length, coding conventions, or clustering choices.

major comments (1)
  1. [Abstract (three-step procedure)] Abstract, description of the three-step procedure: no cluster-quality metrics, linkage/distance sensitivity tests, permutation-based null distributions, or stability checks on the number of clusters are described. This is load-bearing for the central claim, as the seven groups and their 'unique order' Markov chains could reflect the most frequent event multisets, variable follow-up lengths, or arbitrary cluster count rather than genuine temporal interdependence.
minor comments (1)
  1. [Abstract] Abstract: the total sample size (13,192 patients) and AE count (58,575) appear only after the first sentence; moving these figures earlier would improve readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on our exploratory analysis of sequential adverse event patterns following LVAD implantation. We address the concern regarding validation of the clustering step below.

read point-by-point responses

  1. Referee: Abstract, description of the three-step procedure: no cluster-quality metrics, linkage/distance sensitivity tests, permutation-based null distributions, or stability checks on the number of clusters are described. This is load-bearing for the central claim, as the seven groups and their 'unique order' Markov chains could reflect the most frequent event multisets, variable follow-up lengths, or arbitrary cluster count rather than genuine temporal interdependence.

    Authors: We agree that the abstract does not describe these validation elements. The manuscript is explicitly exploratory and selected seven clusters based on the dendrogram combined with clinical interpretability and association with outcomes; however, the referee is correct that additional quantitative checks are needed to rule out artifacts from follow-up duration, event frequency, or arbitrary k. In the revised manuscript we will add: (i) silhouette scores and other cluster-quality metrics, (ii) sensitivity analyses across linkage methods and distance measures, (iii) bootstrap stability assessment of the seven-cluster solution, and (iv) permutation-based null distributions that preserve marginal AE frequencies and censoring patterns to test whether the observed groupings and Markov chains are stronger than expected by chance. These results will be reported in the Methods and Results sections and the abstract will be updated to note the validation steps performed. revision: yes

Circularity Check

0 steps flagged

No circularity: unsupervised sequence mining on external registry

full rationale

The paper applies a standard three-step unsupervised pipeline (sequence construction from INTERMACS registry, hierarchical clustering, per-cluster Markov modeling) to discover seven AE groups. No equations, fitted parameters, or self-citations are shown that reduce the output groups or their temporal orders to quantities defined inside the same model. The derivation is self-contained against the external data source and does not invoke uniqueness theorems or ansatzes from prior author work.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the completeness and chronological accuracy of the INTERMACS adverse-event records and on the validity of the chosen clustering-plus-Markov pipeline for revealing predictive rather than spurious sequences.

free parameters (1)
  • number of clusters
    The procedure groups patients into seven clusters; the choice of seven is presented as the result of the analysis but functions as a modeling decision that shapes the reported groups.
axioms (1)
  • domain assumption Adverse events in the INTERMACS registry are recorded with sufficient completeness and accurate timestamps to support chronological sequencing.
    Step (1) of the procedure converts each patient's AEs into a single chronologically sequenced record; any systematic under-recording or timestamp error would propagate into the clusters.

pith-pipeline@v0.9.0 · 5895 in / 1290 out tokens · 16555 ms · 2026-05-24T20:17:40.706654+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

  1. [1]

    [Online]

    Thoratec, [Online; accessed 11-July-2018]. [Online]. Available: http: //www.thoratec.com/about-us/media-room/library.aspx

    work page 2018

  2. [2]

    National decline in donor heart utilization with regional variability: 1995–2010,

    K. Khush, J. Zaroff, J. Nguyen et al., “National decline in donor heart utilization with regional variability: 1995–2010,” Am. J. Transplantn. , vol. 15, no. 3, pp. 642–649, 2015

    work page 1995

  3. [3]

    Eighth annual INTERMACS report: Special focus on framing the impact of adverse events,

    J. K. Kirklin, F. D. Pagani, R. L. Kormos et al. , “Eighth annual INTERMACS report: Special focus on framing the impact of adverse events,” J. Heart Lung Transplant., vol. 36, no. 10, pp. 1080–1086, 2017

    work page 2017

  4. [4]

    Paving the cowpath: Learning and visualizing clinical pathways from electronic health record data,

    Y . Zhang, R. Padman, and N. Patel, “Paving the cowpath: Learning and visualizing clinical pathways from electronic health record data,” J. Biomed. Inform., vol. 58, pp. 186–197, 2015

    work page 2015

  5. [5]

    Advanced heart failure treated with continuous-flow left ventricular assist device,

    M. S. Slaughter, J. G. Rogers, C. A. Milano et al. , “Advanced heart failure treated with continuous-flow left ventricular assist device,” N. Engl. J. Med. , vol. 361, no. 23, pp. 2241–2251, 2009

    work page 2009

  6. [6]

    Risk factors and outcomes of gastrointestinal bleeding in left ventricular assist device recipients,

    P. S. Joy, G. Kumar, A. K. Guddati et al., “Risk factors and outcomes of gastrointestinal bleeding in left ventricular assist device recipients,” Am. J. Cardiol. , vol. 117, no. 2, pp. 240–244, 2016

    work page 2016

  7. [7]

    GI bleeding in patients with continuous-flow left ventricular assist devices: a systematic review and meta-analysis,

    K. V . Draper, R. J. Huang, and L. B. Gerson, “GI bleeding in patients with continuous-flow left ventricular assist devices: a systematic review and meta-analysis,” Gastrointest. Endosc., vol. 80, no. 3, pp. 435–446, 2014

    work page 2014

  8. [8]

    Adverse events in contemporary continuous-flow left ventricular assist devices: A multi- institutional comparison shows significant differences,

    J. M. Stulak, M. E. Davis, N. Haglund et al. , “Adverse events in contemporary continuous-flow left ventricular assist devices: A multi- institutional comparison shows significant differences,” J. Thorac. Car- diovasc. Surg., vol. 151, no. 1, pp. 177–189, 2016

    work page 2016

  9. [9]

    Major bleeding during HeartMate II support,

    M. C. Bunte, E. H. Blackstone, L. Thuita et al., “Major bleeding during HeartMate II support,” JACC, vol. 62, no. 23, pp. 2188–2196, 2013

    work page 2013

  10. [10]

    Infectious complications in patients with left ventricular assist device: etiology and outcomes in the continuous-flow era,

    V . K. Topkara, S. Kondareddy, F. Malik et al., “Infectious complications in patients with left ventricular assist device: etiology and outcomes in the continuous-flow era,” Ann. Thorac. Surg. , vol. 90, no. 4, pp. 1270– 1277, 2010

    work page 2010

  11. [11]

    Stroke after left ventricular assist device implantation: outcomes in the continuous-flow era,

    L. Harvey, C. Holley, S. S. Roy et al., “Stroke after left ventricular assist device implantation: outcomes in the continuous-flow era,” Ann. Thorac. Surg., vol. 100, no. 2, pp. 535–541, 2015

    work page 2015

  12. [12]

    Outcomes after stroke complicating left ventricular assist device,

    J. Z. Willey, M. V . Gavalas, P. N. Trinh et al., “Outcomes after stroke complicating left ventricular assist device,” J. Heart Lung Transplant. , vol. 35, no. 8, pp. 1003–1009, 2016

    work page 2016

  13. [13]

    Review article: gastrointesti- nal angiodysplasia–pathogenesis, diagnosis and management,

    S. Sami, S. Al-Araji, and K. Ragunath, “Review article: gastrointesti- nal angiodysplasia–pathogenesis, diagnosis and management,” Aliment. pharmacol. & Ther., vol. 39, no. 1, pp. 15–34, 2014

    work page 2014

  14. [14]

    Pre-operative risk factors of bleeding and stroke during left ventricular assist device support: an analysis of more than 900 heartmate ii outpatients,

    A. J. Boyle, U. P. Jorde, B. Sun et al. , “Pre-operative risk factors of bleeding and stroke during left ventricular assist device support: an analysis of more than 900 heartmate ii outpatients,” JACC, vol. 63, no. 9, pp. 880–888, 2014

    work page 2014

  15. [15]

    Risk analysis of bloodstream infection during long-term left ventricular assist device support,

    K. Toda, Y . Yonemoto, T. Fujita et al. , “Risk analysis of bloodstream infection during long-term left ventricular assist device support,” Ann. Thorac. Surg., vol. 94, no. 5, pp. 1387–1393, 2012

    work page 2012

  16. [16]

    Continuous-flow devices and percutaneous site infections: clinical outcomes,

    D. J. Goldstein, D. Naftel, W. Holman et al., “Continuous-flow devices and percutaneous site infections: clinical outcomes,” J. Heart Lung Transplant., vol. 31, no. 11, pp. 1151–1157, 2012

    work page 2012

  17. [17]

    Infectious complications after pulsatile-flow and continuous-flow left ventricular assist device implantation,

    J. M. Schaffer, J. G. Allen, E. S. Weiss et al., “Infectious complications after pulsatile-flow and continuous-flow left ventricular assist device implantation,” J. Heart Lung Transplant. , vol. 30, no. 2, pp. 164–174, 2011

    work page 2011

  18. [18]

    Driveline infections in left ventricular assist devices: implications for destination therapy,

    V . Sharma, S. V . Deo, J. M. Stulak, L. A. Durham et al. , “Driveline infections in left ventricular assist devices: implications for destination therapy,” Ann. Thorac. Surg. , vol. 94, no. 5, pp. 1381–1386, 2012

    work page 2012

  19. [19]

    Neuroendovascular interventions for acute ischemic strokes in patients supported with left ventricular assist devices: A single-center case series and review of the literature,

    F. Al-Mufti, A. Bauerschmidt, J. Claassen et al. , “Neuroendovascular interventions for acute ischemic strokes in patients supported with left ventricular assist devices: A single-center case series and review of the literature,” World Neurosurg, vol. 88, pp. 199–204, 2016

    work page 2016

  20. [20]

    Pre-operative and post-operative risk factors associated with neurologic complications in patients with advanced heart failure supported by a left ventricular assist device,

    T. S. Kato, P. C. Schulze, J. Yang et al., “Pre-operative and post-operative risk factors associated with neurologic complications in patients with advanced heart failure supported by a left ventricular assist device,” J Heart Lung Transplant , vol. 31, no. 1, pp. 1–8, 2012

    work page 2012

  21. [21]

    Predicting post lvad acute severe right heart failure using bayesian analysis,

    M. Kanwar, C. Khoo, L. Lohmueller et al. , “Predicting post lvad acute severe right heart failure using bayesian analysis,” J. Heart Lung Transplant, vol. 38, no. 4, p. S357, 2019

    work page 2019

  22. [22]

    Bayesian model for predicting 90 day event free survival in lvad patients,

    M. Kanwar, L. Lohmueller, R. Kormos et al. , “Bayesian model for predicting 90 day event free survival in lvad patients,” J. Heart Lung Transplant, vol. 37, no. 4, p. S193, 2018

    work page 2018

  23. [23]

    A decision support system for the treatment of patients with ventricular assist device support,

    E. Karvounis, M. Tsipouras, A. Tzallas et al., “A decision support system for the treatment of patients with ventricular assist device support,” Methods Inf. Med. , vol. 53, no. 02, pp. 121–136, 2014

    work page 2014

  24. [24]

    Bleeding and thrombosis associated with ventricular assist device therapy,

    P. Shah, U. S. Tantry, K. P. Bliden, and P. A. Gurbel, “Bleeding and thrombosis associated with ventricular assist device therapy,” J. Heart Lung Transplant, vol. 36, no. 11, pp. 1164–1173, 2017

    work page 2017

  25. [25]

    Process mining in healthcare: A literature review,

    E. Rojas, J. Munoz-Gama, M. Sep ´ulveda et al. , “Process mining in healthcare: A literature review,” J. Biomed. Inform. , vol. 61, pp. 224– 236, 2016

    work page 2016

  26. [26]

    Process mining for clinical pathway: Literature review and future directions,

    W. Yang and Q. Su, “Process mining for clinical pathway: Literature review and future directions,” in 11th Int Conf ICSSSM . IEEE, 2014, pp. 1–5

    work page 2014

  27. [27]

    Process mining in healthcare: A systema- tised literature review,

    M. Ghasemi and D. Amyot, “Process mining in healthcare: A systema- tised literature review,” Int J Electron Healthc , vol. 9, no. 1, pp. 60–88, 2016

    work page 2016

  28. [28]

    Application of process mining in healthcare–a case study in a dutch hospital,

    R. S. Mans, M. Schonenberg, M. Song et al. , “Application of process mining in healthcare–a case study in a dutch hospital,” in Int.l Conf. Biomed. Eng. Syst. Technol. Springer, 2008, pp. 425–438

    work page 2008

  29. [29]

    On process mining in health care,

    U. Kaymak, R. Mans, T. van de Steeg et al. , “On process mining in health care,” in IEEE Int. Conf. Syst. Man Cybern. , 2012, pp. 1859– 1864

    work page 2012

  30. [30]

    Process mining based modeling and analysis of workflows in clinical care-a case study in a chicago outpatient clinic,

    Z. Zhou, Y . Wang, and L. Li, “Process mining based modeling and analysis of workflows in clinical care-a case study in a chicago outpatient clinic,” in IEEE 11th Int. Conf. Netw. Sens. Control, 2014, pp. 590–595

    work page 2014

  31. [31]

    Evaluation of discovered clinical pathways using process mining and joint agent-based discrete-event simulation,

    V . Augusto, X. Xie, M. Prodel et al., “Evaluation of discovered clinical pathways using process mining and joint agent-based discrete-event simulation,” in Proc. Winter Simulation Conf. IEEE Press, 2016, pp. 2135–2146

    work page 2016

  32. [32]

    Temporal abstrac- tions to enrich activity-based process mining corpus with clinical time series,

    C. Fernandez-Llatas, L. Sacchi, J. M. Benedi et al., “Temporal abstrac- tions to enrich activity-based process mining corpus with clinical time series,” in Int. Conf. Biomed. Health Informat. IEEE, 2014, pp. 785– 788

    work page 2014

  33. [33]

    Systematic mapping of process mining studies in healthcare,

    T. G. Erdogan and A. Tarhan, “Systematic mapping of process mining studies in healthcare,” IEEE ACCESS, vol. 6, pp. 24 543–24 567, 2018

    work page 2018

  34. [34]

    On clinical pathway discovery from electronic health record data,

    Y . Zhang, R. Padman, L. Wasserman et al. , “On clinical pathway discovery from electronic health record data,” IEEE Intell. Syst., vol. 30, no. 1, pp. 70–75, 2015

    work page 2015

  35. [35]

    Gender and heart failure: a population perspective,

    P. Mehta and M. Cowie, “Gender and heart failure: a population perspective,” Heart, vol. 92, no. suppl 3, pp. iii14–iii18, 2006

    work page 2006

  36. [36]

    Ward’s hierarchical agglomerative clus- tering method: which algorithms implement ward’s criterion?

    F. Murtagh and P. Legendre, “Ward’s hierarchical agglomerative clus- tering method: which algorithms implement ward’s criterion?” J. clas- sification, vol. 31, no. 3, pp. 274–295, 2014

    work page 2014

  37. [37]

    Algorithmes et applications pour l’extraction de motifs s´equentiels dans le domaine de la fouille de donn´ees: de l’incr´emental au temps r ´eel,

    F. Masseglia, “Algorithmes et applications pour l’extraction de motifs s´equentiels dans le domaine de la fouille de donn´ees: de l’incr´emental au temps r ´eel,” These de doctorat, Universit ´e de Versailles Saint-Quentin- en-Yvelines, 2002

    work page 2002

  38. [38]

    Exploratory mining of life event histories,

    G. Ritschard, R. B ¨urgin, and M. Studer, “Exploratory mining of life event histories,” in Contemporary issues in exploratory data mining in the behavioral sciences. Routledge New York, NY , 2013, pp. 221–253

    work page 2013

  39. [39]

    Markov models in medical decision making: a practical guide,

    F. A. Sonnenberg and J. R. Beck, “Markov models in medical decision making: a practical guide,” Med. Decis. Making , vol. 13, no. 4, pp. 322–338, 1993

    work page 1993

  40. [40]

    Left ventricular assist device performance with long-term circulatory support: lessons from the rematch trial,

    W. P. Dembitsky, A. J. Tector, S. Park et al. , “Left ventricular assist device performance with long-term circulatory support: lessons from the rematch trial,” Ann Thorac Surg , vol. 78, no. 6, pp. 2123–2130, 2004

    work page 2004

  41. [41]

    Ethical challenges with the left ventricular assist device as a destination therapy,

    A. G. Rizzieri, J. L. Verheijde, M. Y . Rady et al. , “Ethical challenges with the left ventricular assist device as a destination therapy,” Philos Ethics Humanit Med , vol. 3, no. 1, p. 20, 2008

    work page 2008

  42. [42]

    Seventh INTERMACS annual report: 15,000 patients and counting,

    J. K. Kirklin, D. C. Naftel, F. D. Pagani, R. L. Kormos et al., “Seventh INTERMACS annual report: 15,000 patients and counting,” J. Heart Lung Transplant., vol. 34, no. 12, pp. 1495–1504, 2015

    work page 2015

  43. [43]

    Sequential pattern mining using a bitmap representation,

    J. Ayres, J. Flannick, J. Gehrke et al., “Sequential pattern mining using a bitmap representation,” in 8th ACM SIGKDD. ACM, 2002, pp. 429– 435

    work page 2002