MRI Driven Diagnostic Management Pathway - Task 1

Purpose

Detection, Anatomical Localization, and Scoring Likelihood of Clinically Significant Prostate Cancer on MRI

Tag(s)

 

Panel

 Abdominal

Define-AI ID

 22020022

Originator

 Martha Menchaca, Katarynza Macura
Lead  Martha Menchaca

Panel Chair

  Luther B. Adair II 

Panel Reviewers

  Luther B. Adair II, Prostate Imaging Subcommittee

License

  Creative Commons 4.0
Status

 Public Comment
RadElement Set   RDES226  
                               

Clinical Implementation

Value Proposition

Multiparametric magnetic resonance imaging (mpMRI) of the prostate is well established as a diagnostic tool for detection and localization of prostate cancer, allowing for targeted sampling of MRI-visible lesions that meet imaging criteria for clinically significant cancer. Known limitations of mpMRI include dependence on optimized imaging quality and experience of the interpreting radiologist, as well as false negativity in MR-visible prostate cancers and false positivity in mimickers of cancer. Thus, AI-driven diagnostic tool for automated detection and classification of prostatic lesions on mpMRI is needed to address the following challenges: 1) variable quality of mpMRI and quality-dependent degradation of imaging, 2) high inter-observer variability of human readers in the interpretation of mpMRI, 3) under-utilization of imaging data not perceptible by the human eye that could enhance detection capability of mpMRI. Successful delivery of the MRI-directed pathway for men with elevated PSA suspected of having prostate cancer relies on maximization of diagnostic capability of mpMRI. 

 

Narrative(s)

58-year-old man with PSA 3.0 ng/mL increasing to 4.0 ng/mL over 12 months, negative DRE, biopsy naïve. 

Workflow Description

The relevant images are obtained from the modality and sent to PAS and the AI engine based on anatomic landmarks. AI-driven automated imaging quality assessment tool processes T2-weighted imaging series and diffusion weighted imaging series to automatically screen for non-diagnostic images. 

Output: Scan flagged as diagnostic vs. non-diagnostic 

Actions: Non-diagnostic scan -> Notification system to alert MRI technologist that imaging quality is non-diagnostic, repeat scan advised; Diagnostic scan -> Acceptance of scan for automated lesion detection task.

Future development: Image quality metric PI-QUAL may be used to score prostate imaging quality rather than using a binary system. AI-driven automated imaging quality assessment tool will detect 90% of non-diagnostic scans.

Considerations for Dataset Development

Procedures

{Multiparametric MRI (DW) Prostate with contrast (DCE MRI), without contrast:}

View(s)

{Basic Parameters}

  • Slice thickness: 3-4 mm without gap

  • Field of View (FOV): 12-20 cm covering the entire prostate and seminal vesicles

  • In plane dimension: <0.7 mm (phase) x <0.4 mm (frequency)


{DW MRI Specifications}

  • Echo Time (TE) < 90 msec; Repetition time (TR) > 3,000 msec

  • Slice Thickness: < 4mm without gap

  • FOV: 16-22 cm covering entire prostate and seminal vesicles

  • In-plane dimensions: <2.5 mm (phase and frequency)


{DCE MRI}

  • TR/TE: <100 mxed/ <5 sec

  • Slice Thickness: 3 mm without gap

  • FOV: 12-20 cm covering the entire prostate and seminal vesicles

  • In plane dimension: <2 mm (phase and frequency)

  • Temporal resolution: <10 sec (<7 sec is referred)

  • Total scanning time: >2 min

  • GBCA dose: 0.1mmol/kg, injection rate of 2-3 cc/sec

Age

[0,90]

Sex at Birth

{Male}

  • PSA

  • RACE

  • FAMILY HISTORY

  • GENOMIC PROFILE

Technical Specifications

Inputs

 

DICOM Study

Procedure

Multiparametric MRI (DW) Prostate with contrast (DCE MRI), without contrast:

Views

Data Type

DICOM

Modality

MRI

Body Region

Prostate

Anatomic Focus

Prostate



Primary Outputs


Detection of Diagnostic versus Non-Diagnostic

RadElement ID

RDE1464

Definition

Detection of Diagnostic versus Non-Diagnostic

Data Type

Categorical

Value Set

  • Diagnostic

  • Non-Diagnostic

Units

N/A

Future Development Ideas

Follow-up with institutional trials, open data science access, and distribution consideration. Future development: image quality metric PI-QUAL may be used to score prostate imaging quality rather than using a binary system.