Resolution and signal-to-noise ratio (SNR)
MRI collects data voxel by voxel. The signal-to-noise ratio (SNR) directly depends on the voxel volume: the smaller the voxel, the fewer hydrogen nuclei it contains, and consequently, the weaker the generated MR signal. This issue is particularly relevant in pediatrics, where small anatomical structures require small voxels. If the field of view (FOV) is reduced without changing the data acquisition matrix, the SNR inevitably drops.
To compensate for image quality loss at low SNR, technical approaches are used:
- Increasing the number of acquisitions (number of acquisitions).
- Decreasing the receiver bandwidth (receiver bandwidth).
- Lowering the parallel imaging factor (parallel imaging factor).
- Applying k-space oversampling (oversampling).
A low SNR value can be accepted if image tissue contrast is optimized to the maximum.
Tissue contrast and pulse sequences
The basic principle of creating high contrast is combining various weightings (T1, T2, proton density) and applying intravenous gadolinium contrast enhancement. The method allows assessing both anatomy and physiological parameters within a single scan.
- T1-weighted images (T1WI): can be performed with or without fat suppression (fat suppression). Fat suppression helps clearly visualize parenchymal organs and enhances the contrast effect. Images without fat suppression are needed to look for fat-containing tumors (e.g., teratomas) and to assess anatomical planes.
- Diffusion-weighted imaging (DWI): captures the restriction of water molecule diffusion. This is an essential tool for examining mass lesions, acting as a surrogate marker for tumor cellularity (cellularity).
- Dynamic contrast-enhanced MRI (DCE-MRI): allows the assessment of tissue perfusion and vascular permeability (an angiogenesis marker).
- MR spectroscopy (MRS): evaluates the biochemical composition of tissues. For example, an elevated choline-to-creatine ratio and a reduced N-acetylaspartate indicate a high-grade tumor.
Scanning planes and artifact management
In modern volumetric (3D) scanning, the initially selected plane is not critically important. The optimal slice is formed during post-processing multiplanar reconstruction (MPR). The evolution of isotropic scanning led to the introduction of volumetric spin-echo sequences (SPACE, VISTA, CUBE) with variable flip angles (variable flip-angle). This elegantly solved the problem of contrast blurring caused by long echo trains in classical techniques.
To eliminate motion artifacts, specialized algorithms are used (PROPELLER, BLADE, RADAR), which distribute noise in radial directions, along with reliable respiratory gating (respiratory gating). This has significantly expanded the application of MRI for assessing lung parenchyma and airways.
Functional and interventional MRI
A separate important area is functional MRI (fMRI). It allows generating color activation maps of the cortex in response to motor, speech, or visual paradigms. The data are coregistered with anatomical images and combined with MR tractography to evaluate white matter tracts. Resting-state fMRI (resting-state fMRI) also exists, evaluating intrinsic brain activity without tasks, which is relevant for patients unable to follow instructions.
Additionally, interventional MRI (iMRI) is actively used in neurosurgery—performing medical procedures directly under scanner guidance.
Class and exam questions
What determines the MR signal intensity at a specific point?
The signal is acquired voxel by voxel. Its intensity is directly proportional to the voxel volume and the number of hydrogen nuclei contained within it.
Why use T1WI without fat suppression?
The absence of fat suppression is necessary for clear visualization of anatomical planes and the detection of specific fat-containing lesions.
What does diffusion-weighted imaging (DWI) show in oncology?
DWI displays areas with restricted water diffusion. In oncological diagnostics, this serves as a reliable surrogate marker of high tumor cellularity.
What is multiplanar reconstruction (MPR)?
This is post-processing of data after 3D scanning. It allows generating the optimal slice plane even after data acquisition is complete.
What else to study on this topic
- Features of cartilage visualization on MRI in children
- Application of BOLD-MRI to assess tissue hypoxia
- Specifics of contrast (gadolinium) administration and the risk of brain depositions
- Assessment of perfusion-diffusion mismatch in ischemic stroke
- Clinical application of resting-state fMRI
Master the entire topic in the "Diagnostic Radiology" course
This is a summary. The course covers everything tested in classes and exams.
- Complete study guide on "MRI Working Principle" and 2,199 more pages of the subject
- Assignments and quizzes after every topic — just like in a colloquium
- AI tutor answers questions on the topic anytime
- 10 video lessons on the subject
Other topics in the "Additional Topics" section
Prepared by WhiteCoat Pro faculty in accordance with the medical school curriculum. For educational purposes only.