Grainy Images in a Large Patient With Atrial Fibrillation
A 118 kg patient (BMI 41) with atrial fibrillation comes for diagnostic coronary angiography. Ventricular rate fluctuates between 90 and 150 beats/min during the runs. The images are grainy, the distal vessels are lost in the lateral projection, and the lateral view also shows cardiac motion blur. The flat detector is positioned about 30 cm above the patient's chest, the field of view covers the whole chest, and no collimation is used. All projections are acquired at 15 frames/s cine with the default pulse width. The technologist proposes repeating every run at 30 frames/s to "freeze the motion" and to brighten the images.
1. Identify the technical causes of the poor image quality and rank the corrections you would make first.
Read the worked answer
Several independent problems are stacked here, and frame rate is the least of them.
- Detector-patient distance of 30 cm: this magnifies the heart, so a large field is needed and geometric unsharpness increases. The detector should be brought as close to the chest as anatomy allows.
- No collimation and full-chest field of view: unnecessary scatter reaches the detector and degrades contrast, especially in a large patient where scatter is already high. Collimate to the heart.
- Penetration/graininess: in a 118 kg patient, the default kVp may be insufficient. Raising kVp and adding a wedge or boresight filter improves penetration and evens out the bright/underexposed regions, at some cost in contrast.
- Motion blur from atrial fibrillation: blur is set by the pulse width of each exposure, not by the number of frames. Shortening the pulse (for example to about 5 ms) is what reduces blur; increasing frames/s only samples the blur more often.
Order of correction: patient-detector geometry, collimation, exposure factors/pulse width - then frame rate.
2. Should every run be repeated at 30 frames/s? Justify your answer using dose and diagnostic yield.
Read the worked answer
No. Doubling the frame rate roughly doubles cine dose for the same exposure per frame, and there is no evidence that 30 frames/s improves diagnostic accuracy for a diagnostic angiogram.
Reasoning:
- Diagnostic runs are static anatomy questions; 15 frames/s is standard and adequate, and motion blur here is a pulse-width problem, not a sampling problem.
- 30 frames/s is justified during PCI - wire, balloon and stent movement, and for fast-moving structures where temporal resolution genuinely matters.
- Atrial fibrillation is a special case: the rhythm is irregular, so ECG-gated triggering will not reliably produce a quiescent window. Short pulse width plus accepting a slightly larger number of runs is more useful than more frames per second.
- Use last-image hold and stored fluoroscopy rather than additional cine runs for review, and reduce pulse rate during fluoroscopy to the lowest setting that lets you see wire movement.
The judgment is: spend dose on the structure that needs temporal resolution, not on every projection by default.
3. How does body habitus change your choices of kVp, focal spot and magnification, and what trade-offs do you accept?
Read the worked answer
In a large patient the goals conflict: you need penetration and enough photons, but you also need contrast and sharpness.
- kVp: raising kVp increases penetration and detector signal, reducing graininess and patient entrance dose, but lowers subject contrast. In this patient the net effect is favorable, because the current images are photon-starved rather than contrast-limited.
- Focal spot: the small focal spot may not tolerate the tube current required; you may be forced to the large spot, which increases geometric unsharpness. Compensate by minimizing object-detector distance and avoiding magnification.
- Magnification: avoid it unless a small structure must be resolved. Magnification roughly doubles dose for the same image and enlarges the penumbra; if it is needed, apply it only to the specific run.
- Grid: in very large patients, grid removal reduces dose and increases signal but lets scatter through, which can cost more contrast than it gains; it is a selective option, not a default.
Overall: fix geometry first - those choices cost nothing in dose and improve sharpness at the same time as reducing the need for magnification.