Consiglio Nazionale delle Ricerche

Tipo di prodottoArticolo in rivista
TitoloConsidering patient clinical history impacts performance of machine learning models in predicting course of multiple sclerosis
Anno di pubblicazione2020
FormatoElettronico
Autore/iSeccia R.; Gammelli D.; Dominici F.; Romano S.; Landi A.C.; Salvetti M.; Tacchella A.; Zaccaria A.; Crisanti A.; Grassi F.; Palagi L.
Affiliazioni autoriDept. of Computer, Control and Management Engineering Antonio Ruberti, Sapienza University of Rome, Rome, Dept. of Computer, Control and Management Engineering Antonio Ruberti, Sapienza University of Rome, Rome, Italy; Dept. of Neurosciences, Mental Health and Sensory Organs, Sapienza University of Rome, Rome, Dept. of Neurosciences, Mental Health and Sensory Organs, Sapienza University of Rome, Rome, Italy; IRCCS Istituto Neurologico Mediterraneo Neuromed, Pozzilli, IRCCS Istituto Neurologico Mediterraneo Neuromed, Pozzilli, Italy; Dept. of Physics, Istituto dei Sistemi Complessi (ISC)-CNR, UOS Sapienza, Rome, Dept. of Physics, Istituto dei Sistemi Complessi (ISC)-CNR, UOS Sapienza, Rome, Italy; Dept. of Physics, Sapienza University of Rome, Rome, Dept. of Physics, Sapienza University of Rome, Rome, Italy; Dept. of Physiology and Pharmacology, Sapienza University of Rome, Rome, Dept. of Physiology and Pharmacology, Sapienza University of Rome, Rome, Italy; Technical University of Denmark, Kongens-Lyngby, Current address: Technical University of Denmark, Kongens-Lyngby, Denmark
Autori CNR e affiliazioni
  • ANDREA TACCHELLA
  • ANDREA ZACCARIA
Lingua/e
  • inglese
AbstractMultiple Sclerosis (MS) progresses at an unpredictable rate, but predictions on the disease course in each patient would be extremely useful to tailor therapy to the individual needs. We explore different machine learning (ML) approaches to predict whether a patient will shift from the initial Relapsing-Remitting (RR) to the Secondary Progressive (SP) form of the disease, using only "real world" data available in clinical routine. The clinical records of 1624 outpatients (207 in the SP phase) attending the MS service of Sant'Andrea hospital, Rome, Italy, were used. Predictions at 180, 360 or 720 days from the last visit were obtained considering either the data of the last available visit (Visit-Oriented setting), comparing four classical ML methods (Random Forest, Support Vector Machine, K-Nearest Neighbours and AdaBoost) or the whole clinical history of each patient (History-Oriented setting), using a Recurrent Neural Network model, specifically designed for historical data. Missing values were handled by removing either all clinical records presenting at least one missing parameter (Feature-saving approach) or the 3 clinical parameters which contained missing values (Record-saving approach). The performances of the classifiers were rated using common indicators, such as Recall (or Sensitivity) and Precision (or Positive predictive value). In the visit-oriented setting, the Record-saving approach yielded Recall values from 70% to 100%, but low Precision (5% to 10%), which however increased to 50% when considering only predictions for which the model returned a probability above a given "confidence threshold". For the History-oriented setting, both indicators increased as prediction time lengthened, reaching values of 67% (Recall) and 42% (Precision) at 720 days. We show how "real world" data can be effectively used to forecast the evolution of MS, leading to high Recall values and propose innovative approaches to improve Precision towards clinically useful values.
Lingua abstractinglese
Altro abstract-
Lingua altro abstract-
Pagine da0230219-1
Pagine a0230219-18
Pagine totali-
RivistaPloS one
Attiva dal 2006
Editore: Public Library of Science - San Francisco, CA
Paese di pubblicazione: Stati Uniti d'America
Lingua: inglese
ISSN: 1932-6203
Titolo chiave: PloS one
Titolo proprio: PloS one
Titolo abbreviato: PLoS ONE
Titoli alternativi:
  • Public Library of Science one
  • PLoS 1
Numero volume della rivista15
Fascicolo della rivista3
DOI10.1371/journal.pone.0230219
Verificato da referee-
Stato della pubblicazionePublished version
Indicizzazione (in banche dati controllate)
  • Scopus (Codice:2-s2.0-85082116501)
  • ISI Web of Science (WOS) (Codice:000535303100021)
Parole chiaveArticle, human, machine learning, medical history, multiple sclerosis, outpatient
Link (URL, URI)https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0230219
Titolo parallelo-
Licenza-
Scadenza embargo-
Data di accettazione24/02/2020
Note/Altre informazioni-
Strutture CNR
  • ISC — Istituto dei sistemi complessi
Moduli/Attività/Sottoprogetti CNR-
Progetti Europei-
Allegati