Cardiomegaly, Consolidation, Edema Lung, Lung Injury, Pleural Effusion, Pneumothorax
Conditions
Keywords
Artificial Intelligence, Chest X-ray, Stepped wedge trial
Brief summary
This study aims to evaluate whether the use of AI as a physician support tool is associated with an increase in the detection rate of chest radiographic findings in adults with respiratory complaints, compared to diagnosis performed exclusively by doctors, without AI support. This is a cluster-randomized clinical trial, following the stepped wedge design, and adhering to the guidelines of the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT). In this study, the Diagnostic Support Solution for Chest X-rays - LungAnalysis (LuAna), developed by the Hospital Israelita Albert Einstein (HIAE) within the PROADI-SUS Banco de Imagens, was used. The clinical trial will be conducted in multiple centers with a diverse population from the public health system, to ensure that the algorithms are validated across a broad demographic profile. The expected benefits are significant, providing greater security for patients, increasing doctors' confidence in interpreting chest X-rays, promoting efficiency and cost savings for healthcare services, and offering promising prospects for other AI applications in imaging diagnostics.
Detailed description
Imaging diagnostic aid tools that use AI and facilitate the identification of findings on chest x-rays can contribute to doctors' care routines and clinicians' and radiologists' reporting routines, as these tools can allow the organization of care queues according to priorities, in addition to identifying subtle findings on the image, thereby reducing errors in reading the RXT and benefiting patients with greater agility in care and a shorter time until diagnosis. However, for reliability, these tools must undergo rigorous validation processes in large populations before implementation.
Interventions
DEVICEApp LuAna
Inclusion of chest x-ray images in the LuAna app to receive feedback on lung findings.
Sponsors
Hospital Israelita Albert Einstein
Study design
Allocation
NA
Intervention model
SINGLE_GROUP
Primary purpose
OTHER
Masking
NONE
Intervention model description
This is a type of "stepped wedge" cluster randomized clinical trial. In the model standard of this design, interventions are implemented in the clusters randomly and graduate.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Non-reported chest X-rays (XRts) of individuals aged over 18 years. * Individuals images with respiratory complaints. * Chest X-rays taken during the presence of these respiratory symptoms or while being followed up for respiratory disease. * Chest X-rays taken on any X-ray machine. * Chest X-rays that include at least one frontal view of the chest.
Exclusion criteria
* Those whose chest X-ray was performed due to a history of trauma, pre-operative risk assessment, lung cancer screening, or exclusively for verifying the correct positioning of a peripheral intravenous catheter (PICC). * Chest X-rays with technical quality below the minimum required for proper interpretation and diagnosis. * Cases without at least one frontal view. * X-rays printed on regular paper.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Detection rate | through study completion, an average of 1 year | Detection rate of "Radiological Findings", before and after Artificial Intelligence assistance, compared to gold standard (report validated twice by thoracic radiologists blind to the interpretation of the examining physician and AI result). |
Countries
Brazil
Contacts
CONTACTJoselisa Paiva, PhD
PRINCIPAL_INVESTIGATORJoselisa Paiva, PhD
Hospital Israelita Albert Einstein
Outcome results
None listed