Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
| Name | Class |
|---|---|
| Pius Brinzeu Timisoara County Emergency Hospital | UNKNOWN |
Not provided
Not provided
Not provided
Not provided
The aim of the study is to asses the influence of computer aided diagnostic to the process of lung affection quantification on computer tomography in COVID-19 confirmed patients.
The lung involvement of COVID-19 patients has been showed to be correlated to clinical outcomes and became part of the clinical practice. Even though various scores can be used, the affection estimation is usually done on computer tomography, using radiologists's estimation skills which is a highly subjective process.
Artificial intelligence is a known objective constant and therefore a potential radiologist complement. This trial aims at studying the effect of using a computer aided diagnostic software integrated in the normal clinical practice of radiologists from Timisoara County Emergency Hospital. It uses the AI-PROBE analysis setup, which turns off the CAD outputs for randomly chosen 50% the cases (control) and then compares the radiological reports for differences between the two arms.
Not provided
Not provided
Not provided
Not provided
| Label | Type | Description | Intervention Names |
|---|---|---|---|
| CAD analysis | Experimental | The XVision COVID-19 computer aided diagnostic software is used by radiologist at CT analysis time |
|
| No CAD analysis | No Intervention | No CAD analysis is shown to radiologist. |
| Name | Type | Description | Arm Group Labels | Other Names |
|---|---|---|---|---|
| CAD analysis | Diagnostic Test | CAD shows the radiologist automatically delineated areas of potential COVID-19 affection, together with an overall lung affection percentage. |
|
| Measure | Description | Time Frame |
|---|---|---|
| Mean difference of lung affection quantification percentage | The objective measurement of lung affection percentage is measured against pixel level labels. A lower difference mean better outcome. | At CT acquisition time, up to 2 weeks |
Not provided
Not provided
Inclusion Criteria:
Exclusion Criteria:
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
| Facility | Status | City | State | ZIP | Country | Contacts |
|---|---|---|---|---|---|---|
| Pius Brinzeu Timisoara County Emergency Hospital | Timișoara | Timiș County | Romania |
| PubMed Identifier | Type | Citation | Retractions |
|---|---|---|---|
| 36966179 | Derived | Bercean BA, Birhala A, Ardelean PG, Barbulescu I, Benta MM, Rasadean CD, Costachescu D, Avramescu C, Tenescu A, Iarca S, Buburuzan AS, Marcu M, Birsasteanu F. Evidence of a cognitive bias in the quantification of COVID-19 with CT: an artificial intelligence randomised clinical trial. Sci Rep. 2023 Mar 25;13(1):4887. doi: 10.1038/s41598-023-31910-3. |
Not provided
Not provided
The statistical analysis, data points and possibly deidentified images.
Not provided
Not provided
Not provided
Not provided
Not provided
| ID | Term |
|---|---|
| D000086382 | COVID-19 |
| ID | Term |
|---|---|
| D011024 | Pneumonia, Viral |
| D011014 | Pneumonia |
| D012141 | Respiratory Tract Infections |
| D007239 | Infections |
Not provided
Not provided
CAD outputs are turned off for randomly chosen 50% of patients, which represent the control group. The other 50% are analysed using CAD
Not provided
Not provided
The random assignment is done automatically by the CAD system and is not visible to the patient. The radiologist obviously sees which cases have CAD analysis and which not.
| D014777 |
| Virus Diseases |
| D018352 | Coronavirus Infections |
| D003333 | Coronaviridae Infections |
| D030341 | Nidovirales Infections |
| D012327 | RNA Virus Infections |
| D008171 | Lung Diseases |
| D012140 | Respiratory Tract Diseases |