Identifying Small Bowel Obstruction

Purpose	Detect small bowel obstruction
Tag(s)
Panel	Abdominal Panel
Define-AI ID	20020014
Originator	Zerwa Farooq
Panel Chair	Luther B. Adair
Panel Reviewers	Luther B. Adair, Andrew Smith, and Kiran Thakrar
License	Creative Commons 4.0
Status	Public Comment
RadElement Set	RDES178

Clinical Implementation

Value Proposition

Small bowel obstruction is a common surgical condition accounting for up to 20% of surgical admissions. Evaluation of the presence, severity, cause and complications of small bowel obstruction on cross sectional imaging provides the basis of surgical or conservative management strategies. An algorithm evaluating the above mentioned characteristics can expedite management and aid radiologists in accurately and quickly classifying small bowel obstruction cases.

Narrative(s)

A 50-year-old woman presented with a 2-day history of worsening nausea, vomiting, and constipation. She had a surgical history of appendectomy. An algorithm evaluates the small bowel and informs the radiologist regarding the presence or absence of obstruction.

Workflow Description

Images are obtained from the CT and sent to PACS and the AI engine. The image is then analyzed by the AI engine and an alert is sent to PACS with the classification information. An alert is also sent to the ordering provider so that appropriate consultation and management can be initiated without unnecessary delay.

Considerations for Dataset Development

Procedure	CT, CECT, CT with or without oral contrast
Sex at birth	Male, female
Age	0-90
Presence of SBO	Dilated proximal loops of small bowel with relatively collapsed distal small bowel loops (in contradistinction to ileus where no transition point is identified), extent of dilation of small bowel loops that can be measured as maximum diameter in millimeters
Degree of obstruction	Complete obstruction, high grade or low grade partial obstruction, based on the passage of oral contrast beyond the point of obstruction.
Transition point	High or low transition point
Cause of obstruction	Intrinsic (inflammatory, neoplasm, hematoma), extrinsic (adhesions, hernias, external mass effect) or intraluminal (foreign body, gallstones, bezoars)
Presence of complications	Closed loop obstruction, bowel wall thickening, bowel wall enhancement, pneumatosis, ascites, pneumoperitoneum

Technical Specifications

Inputs

DICOM Study

Procedure	CT
Data Type	DICOM
Modality	CT
Body Region	Abdomen Pelvis
Anatomic Focus	Small bowel

Primary Outputs

Small bowel obstruction

RadElement ID	RDE1237
Definition	Small bowel obstruction
Data Type	Categorical
Value Set	SBO absent SBO present undetermined
Units	N/A

Secondary Outputs

Severity of obstruction

RadElement ID	RDE1238
Definition	Severity of obstruction
Data Type	Categorical
Value Set	partial low grade partial high grade complete obstruction undetermined
Units	N/A

Small bowel diameter

RadElement ID	RDE1239
Definition	Maximum small bowel diameter
Data Type	Numeric
Value Set	0-100
Units	mm

Location of obstruction

RadElement ID	RDE1240
Definition	Location of obstruction
Data Type	Categorical
Value Set	proximal mid distal undetermined
Units

Cause of obstruction

RadElement ID	RDE1241
Definition	Cause of obstruction
Data Type	Categorical
Value Set	intrinsic extrinsic intramural undetermined
Units	N/A

Small bowel wall enhancement

RadElement ID	RDE1242
Definition	Small bowel wall enhancement
Data Type	Categorical
Value Set	diffuse present partial absent completely absent undetermined
Units	N/A

Small bowel wall thickness

RadElement ID	RDE1243
Definition	Small bowel wall thickness
Data Type	Numeric
Value Set	0 - 10
Units	mm

Presence of ascites

RadElement ID	RDE1244
Definition	Ascites
Data Type	Categorical
Value Set	absent present undetermined
Units	N/A

Presence of pneumatosis

RadElement ID	RDE1245
Definition	Pneumatosis
Data Type	Categorical
Value Set	absent present undetermined
Units	N/A

Presence of pneumoperitoneum

RadElement ID	RDE1246
Definition	Pneumoperitoneum
Data Type	Categorical
Value Set	absent present undetermined
Units	N/A

Future Development Ideas

The algorithm can compare studies with prior established SBO and appropriately assess resolution or worsening of obstruction e.g. interim resolution of bowel dilation or development of indicators suggesting complications as explained above.

Related Datasets

No known related public datasets at this time, please alert us if you know of any.