Hepatic Volumes

Purpose

 AI would be beneficial for the evaluation of efficient and accurate hepatic volumes

Tag(s)

Panel

Abdominal Panel

Define-AI ID

21020017

Originator

Jewel Appleton
Lead Jewel Appleton

Panel Chair

Luther B. Adair

Panel Reviewers

Luther B. Adair & Andrew Smith

License

Creative Commons 4.0

Status Public Comment
RadElement Set RDES190

 
   
   
   
   
   
   
   
   
                               

Clinical Implementation


 

Value Proposition


The value this algorithm would add is reproducible anatomic volumes and speed of segmentation in patients with a liver mass undergoing pre-operative surgical planning, in liver donors, and in patients with chronic liver disease undergoing CT imaging for staging of liver fibrosis.


Narrative(s)


50-year-old-male patient with a single 3 cm HCC in the right lobe. Given proximity to the vascular pedicle the surgeon is requesting volume measurements of the right and left lobes before resection.

 

36-year-old male potential liver donor undergoing pre-operative planning.


55-year-old female with chronic hepatitis C virus has elevated liver function tests, and there is a clinical need to undergo staging of hepatic fibrosis.








Workflow Description

 

A patient is identified by the primary surgical team as a candidate for hepatic resection. Arterial and portal venous phase CT images of the liver are uploaded to the AI program. The AI automatically segments the various liver segments. 3D reconstructions are generated. The liver segmental volumes are extracted individually with automated calculations of total liver, right lobe, left lobe, and the segmental volume ratio (segments I-III relative to IV-VIII). A radiologist reviews the segmentations and adjusts / edits as needed. A report is generated and includes the metrics and key images. The volumetric data and report are sent to the PACS. Surgical planning and reassurance is performed using this data to ensure that the patient is an adequate candidate for resection.









Considerations for Dataset Development



Procedures

CT

View(s)

Contrast enhanced series of CT

Age

Pediatric 0-17 years old and Adult 18-75

Confounders

Prior surgical resection, metallic clips, poor image quality, motion artifact, unexpected patient positioning

Technical Specifications



Input


DICOM Study

Procedure

CT

Views

Contrast enhanced series of CT

Data Type

DICOM

Modality

CT

Body Region

Abdomen

Anatomic Focus

Liver




Primary Outputs


Hepatic tumor location 


RadElement ID

RDE1274

Definition

Detect tumor location

Data Type

Categorical

Value Set


  • right upper quadrant

  • right lower quadrant

  • left upper quadrant

  • left lower quadrant


Units

N/A





Secondary Outputs


Volume of entire liver

RadElement ID

RDE1275

Definition

volume of entire liver

Data Type

Numeric

Value Set


Units

ml



Volume of right liver lobe

RadElement ID

RDE1276

Definition

volume of right liver lobe

Data Type

Numeric

Value Set


Units

ml


Volume of left liver lobe

RadElement ID

RDE1277

Definition

volume of left liver lobe

Data Type

Numeric

Value Set


Units

ml



Volume of  tumor 

RadElement ID

RDE1278

Definition

volume of tumor

Data Type

Numeric

Value Set


Units

ml

Future Development Ideas


  • This information can also be used to evaluate tumor burden. 

  • We can also extrapolate the data to follow up cases in the development of fibrosis or cirrhosis.