10.1 Biomarkers in MRI Flashcards
(16 cards)
What are the requirements for an MRI biomarker?
images need to be quantifiable (convert image -> numbers) AND link this number to disease
What is the issue with conventional qualitative MRI?
only image contrast can be meaningfully interpreted (not image intensity)
What is the trade-off with QUANtitative MRI?
you have to compromise between accuracy of the scan and practicality/time scan takes
what is an additional burden that quantitative scans have (compared to qualitative)?
quan: you have to do additional scans to make quan scans -> increases time
AND also more complex analysis
What is the issue with using qualitative MRI to treatment management of multiple sclerosis?
What is a solution to this?
-Multiple sclerosis = brain degenerative disease where patients can progress really rapidly, slowly or medium
With qualitative MRI -> you don’t know how rapid it is and cant gauge how aggressive treatment needs to be
-use quantitative MRI
What is the issue with replicability of qualitative images?
Is this seen in quantitative images?
-different scanners produce different images of the same anatomy
-no! produces same results on any scanner anywhere in the world
Can image intensity be interpreted on qualitative and quantitative MRI scans?
no only image intensity has meaningful on quantitative
Give an example of an MRI biomarker for stroke management?
decide whether patient needs thrombolysis treatment (dissolve blood clots) to prevent further stroke -> mismatch between stroke areas in DWI images and perfusion images (MTT) is quantified -> gives a number which helps clinician decide if patient needs thrombolysis
Give an example of an MRI biomarker in cancer patients?
use CBV CBF to see blood flow metrics of tumour before and after radiotherapy
Which is easier to use for MRI biomarkers: quan or qual images?
quan as you already have the number that you need to link to biology/disease
What is the advantages of using MRI biomarkers over biopsy and biofluids(blood, urine)?
-biopsy is invasive and MRI is not -> MRI biomarker is repeatable
no sampling error in MRI biomarkers as biopsy can be impacted by disease heterogeneity
MRI biomarkers don’t disturb tissue state
-MRI markers more sensitive and specific to origin than biofluids because biofluids can be confounded by changes outside target organ
-biopsy and biofluids and NOT in situ whereas MRI is
What are the advantages of using MRI biomarkers?
-MRI is versatile capturing a multi-scale assessment of an organ (ie can look at haemodynamics, morphology etc)
-MRI can look at multiple organs at the same time to see how they interact in real-time
What is the technical issue with developing new MRI biomarkers for clinical practice?
some MRI biomarkers needs or would benefit from higher strength scanners 7Tesla but few hospitals have these higher power scanners
researchers are developing 11+ Tesla scanners but they are huge!
the high power scanners are feasible for only neuroscience research not everyday clinical practice
What is the interest of the 2010s MRI biomarkers research?
emergence of quantitative radiology caused a shift towards better data instead of trying to get better high power scanners/hardware
What is the benefit of using AI in MRI biomarkers?
-AI tool draws line over area of interest/diseased area instead of having to employ an expert clinician to look over scans
What are some issues with developing MRI biomarkers for clinical practice?
-costly development
-scanners availability in hospitals
-still heavily relies on expert intervention (to draw lines)
