Keywords:
Computer applications, Thorax, Conventional radiography, Education, eLearning, Education and training
Authors:
M. C. Teruel-Garrido, R. Lorenzo-Alvarez, F. Sendra Portero; Malaga/ES
DOI:
10.1594/ecr2018/C-1021
Methods and materials
Radiotorax.es
Radiotorax.es is a Web-based platform created to train the interpretation of chest radiographs.
Users must first register and login.
They have two series of three cases to train and become familiar with the online procedure.
The training consists of reporting series of 20 cases and self-assess their reports against those of an experienced radiologist.
The user must evaluate in each case the following aspects:
- Whether he wrote the report correctly (Yes/No)
- Whether he described all the findings (Yes/No)
- Whether he described a specific requested finding (Yes/No).
In case of normal images,
he was asked if he had described the study as normal
- Indicate the degree of difficulty of the radiography (Very low,
low,
medium,
high or very high)
- Rate his radiological report from 1 (drewful) to 10 (excellent)
At the end,
a PDF file is generated with all the self-assessment information.
Radiotorax.es manages a database of 200 cases,
distributed into normal (35%),
frequent (35%),
subtle (15%) and serious (15%) findings.
Users can conduct self-assessment rounds as many times as they want.
For each round the application provides 20 randomized cases with the same proportions.
On December 1st 2016 there were 5267 users registered.
Evaluation
During five successive years (2012-2016),
third-year medical students were asked to perform four monthly self-assessments using Radiotorax.es during the four-month Radiology course,
and their corresponding PDF-files were collected.
The mean results of each self-assessment were organized in Excel files.
Descriptive statistical values (mean,
standard deviation and variance) were calculated for the total of students of each group and for each item.
Student T-test was used to compare the progressive values within a course (paired data) and between different courses (unpaired data).