“Physit”—A diagnosis and troubleshooting tool for physiotherapists in training

Reuth Mirsky, Shay Hibah, Moshe Hadad, Ariel Gorenstein, Meir Kalech

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Many physiotherapy treatments begin with a diagnosis process. The patient describes symptoms, upon which the physiotherapist decides which tests to perform until a final diagnosis is reached. The relationships between the anatomical components are too complex to keep in mind and the possible actions are abundant. A trainee physiotherapist with little experience naively applies multiple tests to reach the root cause of the symptoms, which is a highly inefficient process. This work proposes to assist students in this challenge by presenting three main contributions: (1) A compilation of the neuromuscular system as components of a system in a Model-Based Diagnosis problem; (2) The PhysIt is an AI-based tool that enables an interactive visualization and diagnosis to assist trainee physiotherapists; and (3) An empirical evaluation that comprehends performance analysis and a user study. The performance analysis is based on evaluation of simulated cases and common scenarios taken from anatomy exams. The user study evaluates the efficacy of the system to assist students in the beginning of the clinical studies. The results show that our system significantly decreases the number of candidate diagnoses, without discarding the correct diagnosis, and that students in their clinical studies find PhysIt helpful in the diagnosis process.

Original languageEnglish
Article number72
JournalDiagnostics
Volume10
Issue number2
DOIs
StatePublished - 1 Jan 2020

Keywords

  • Applications
  • Diagnosis
  • Model based diagnosis
  • Physiotherapy

ASJC Scopus subject areas

  • Clinical Biochemistry

Fingerprint

Dive into the research topics of '“Physit”—A diagnosis and troubleshooting tool for physiotherapists in training'. Together they form a unique fingerprint.

Cite this