Motor Cortex Quantitative Susceptibility Mapping

Purpose

Quantification of QSM of motor cortex and segmentation

Tag(s)

 

Panel

 Neuroradiology

TOUCH-AI ID

 TAI-03180002

Originator

 John Tsiouris

Panel Chair

 Alex Norbash

Panel Reviewers

 Neuroradiology Panel 

License

 Creative Commons 4.0

Status  Public Commenting 
                               

Clinical Implementation


Value Proposition

The diagnosis of primary upper motor neuron diseases such as amyotrophic lateral sclerosis (ALS) and primary lateral sclerosis (PLS) is a clinical challenge. Currently, the role of imaging is to exclude structural lesions that may mimic these diseases. Diagnosis is commonly delayed, and false positive diagnoses can occur. Therefore, there has been recent research interest in improved imaging biomarkers for ALS/PLS. Susceptibility weighted imaging (SWI) with quantitative susceptibility mapping (QSM) has become an intriguing imaging biomarker. Numerous recent publications indicate promise in differentiating patients with a primary upper motor neuron disease from normals (Schweitzer et al, AJR 2015; Adachi et al, Journal of Neuroimaging 2014) and mimics (Lee et al, Neuroradiology 2017). Currently, manual segmentation and QSM assessments of the motor cortex are necessary, difficult, and time consuming. Automating this procedure with machine learning would facilitate research and assist in the development of a promising imaging biomarker.

Narrative(s)

A 56-year-old man is having new difficulty walking and occasionally trips and falls. He is also having progressive difficulty writing and holding his eating utensils. Occasionally, he sees the muscles in his legs twitching involuntarily. These symptoms are concerning, and he sees a neurologist.

Workflow Description

The patient is seen by a neurologist, who suspects a motor neuron disease. He orders MRI scans of the brain and spine to assess for any structural lesions or demyelination.

An algorithm receives a post-processed SWI/QSM data set. If the algorithm can determine a result, return the following: representative images of motor cortex segmentation, quantitative measures of the entire motor cortex (left and right, separately), QSM measurements of the face/hand/leg portions of the motor cortex homunculus, and the odds/risk ratios for ALS/PLS given these results.

Additional considerations are as follows: The algorithm executes after the exam is verified on PACS. The algorithm optimally integrates on PACS and dictation/reporting software. The user is able to automatically populate the report or manually input the results. An indicator image may save to PACS as part of the medical record.

Considerations for Dataset Development


Procedures(s): MRI, Brain, SWI with QSM

Sex at Birth: {Male, Female}

 

Age (years): [21,90]

Technical Specifications


Inputs

DICOM Study

Procedure

MRI, Brain, SWI with QSM

Data Type

DICOM

Modality

MRI

Body Region

Head

Anatomic Focus

Brain

Primary Outputs

Motor Cortex QSM Mean

RadElement ID

RED310

Definition

Mean calculation of QSM MR units of the left and right motor cortex

Data Type

Numeric

Value Set

[−100,100]

Units

MR Units

 

Motor Cortex QSM Max

RadElement ID

RDE311

Definition

Max calculation of QSM MR units of the left and right motor cortex

Data Type

Numeric

Value Set

[−100,100]

Units

MR Units

 

Motor Cortex QSM Standard Deviation

RadElement ID

RDE312

Definition

Standard deviation of QSM MR units of the left and right motor cortex

Data Type

Numeric

Value Set

[−100,100]

Units

MR Units

Future Development Ideas


Display segmentation map of motor cortex into left/right and face/hand/leg components.

Public Commenting


Use cases are meant to be a primary vehicle for distributing clinical information to the developer community. They pinpoint precise scenarios within radiology workflows where potential automation could add noticeable value and establish standards for interpreting and passing corresponding common data elements. Implementing effective standards requires the perspective from all stakeholders. So to that end, we encourage your feedback on use cases.

To submit comments, please email DSIUseCases@acr.org with the use case title(s) and relevant comments by January 1, 2019. If more convenient, you may also download this use case and comment directly on the PDF. Just attach the PDF copy on the email.