|
Purpose |
To assess the position of enteric tube placement. Enteric tubes refer to support devices placed for feeding patients who cannot swallow or for decompressing the GI tract. The tip of these tubes needs to be in the correct location to function, and a misplaced tube that is used can injure the patient. A frequent indication of chest and abdominal radiographs is to check position of the tube and an algorithm that could determine placement would automate this task. |
|
Tag(s) |
|
|
Panel |
Abdominal Panel |
|
Define-AI ID |
20020012 |
|
Originator |
Christopher P. Gange |
| Lead | Christopher P. Gange |
|
Panel Chair |
Luther B. Adair |
|
Panel Reviewers |
Abdominal Panel |
|
License Status RadElement Set |
|
Evaluation of tube placement on radiographs is a high volume, low priority job for radiologists. An algorithm that could automate this process would improve workflow. In addition, an abnormal result from this algorithm could be used as part of a system that elevates the priority of unread studies to lower the turnaround time for abnormal results that affect patient care.
A 65-year-old patient with abdominal pain is found on imaging to have a bowel obstruction and has an oral-gastric tube placed for decompression. After the radiograph to check position is completed, it takes over 45 minutes for a radiologist to look at the images and find that the distal aspect of the tube is not within the stomach. During this time, the patient could have an aspiration event, which could have been prevented if the misplaced tube was identified more quickly.
An 88-year-old patient with failure to thrive is to be started on tube feeds to improve nutrition. A feeding tube placement is attempted but a chest radiograph shows that the tube is malpositioned with the tip located in the right lung. It takes the clinical team 3 more attempts to correctly place this tube and begin refeeding the patient.
The images are obtained from modality and sent to PACS and the AI engine. The image is analyzed by the engine. The system first determines the coordinates of the tip’s location. The system then determines whether the tube is positioned correctly based on the tip location. If an abnormal position is found, an alert message is sent to PACS from the engine with the information about the current study, and graphic highlighting the enteric tube tip.
|
Procedure |
X-Ray, Abdomen/Chest |
|
View(s) |
Chest: AP or PA, Abdominal: supine or upright |
|
Age |
0-100 |
|
Sex at birth |
Male, Female |
|
Comorbidities |
lung disease (e.g., Pneumonia, emphysema, bronchiectasis, malignancy), cardiac disease, abdominal disease (bowel obstruction, ileus, malignancy) recent trauma, recent surgery |
|
Other lines and tubes present |
Central venous catheters, pleural chest tubes, ECMO cannulas, Endotracheal tubes, surgical drains, gastrostomy tubes, nephrostomy tubes pacer devices. |
|
Post-surgical findings/devices |
Spinal hardware, sternotomy wires, valve replacements, surgical clips, spinal stimulator leads, etc. |
DICOM Study
|
Procedure |
X-Ray, Abdomen/Chest |
|
Views |
Chest: AP or PA, Abdominal: supine or upright |
|
Data Type |
DICOM |
|
Modality |
X-Ray |
|
Body Region |
Chest/Abdomen |
|
Anatomic Focus |
Tube Tip Location
|
RadElement ID |
|
|
Definition |
Identify the location of tube’s tip |
|
Data Type |
Coordinates |
|
Value Set |
N/A |
|
Units |
N/A |
Tube Side-port Location
|
RadElement ID |
|
|
Definition |
Identify the location of tube’s side port |
|
Data Type |
Coordinates |
|
Value Set |
N/A |
|
Units |
N/A |
Tube Placement
|
RadElement ID |
|
|
Definition |
Determine if the tube placement is correct |
|
Data Type |
Categorical |
|
Value Set |
|
|
Units |
N/A |
Ability to identify types of tubes.
Ability to take an anatomic location as an input and output whether the location is correct.