Visual Examples
Explore VQA examples under medical image quality degradations. See how different severity levels (L0/L1/L2) affect clinical question answering.
Artifacts
Limited angle, sparse view, bias field, undersampling, ghosting, blood cell, and dark spots artifacts
Bias Field Artifact
MRI GMAIMMbench Organ Recognition - Abdomen
L0 (Original)
L1 (Moderate)
L2 (Severe)
Observe the MRI image. Can you identify the organ in the highlight area?
Blood Cell Artifact
Histopathology GMAIMMbench Cell Recognition
L0 (Original)
L1 (Moderate)
L2 (Severe)
Which cell type is indicated in the highlighted area of the Histopathology image?
Dark Spots Artifact
Histopathology GMAIMMbench Cell Recognition
L0 (Original)
L1 (Moderate)
L2 (Severe)
Which of the following options best matches the marked organ in the Histopathology image?
Ghosting Artifact
MRI GMAIMMbench Organ Recognition - Abdomen
L0 (Original)
L1 (Moderate)
L2 (Severe)
Observe the MRI image. Can you identify the organ in the highlight area?
Limited Angle
CT GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
This is a CT image. Which of the following options is the most appropriate to demonstrate the marked area?
Sparse View
CT GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Considering the symptoms visible in highlighted area of the CT picture, which option corresponds most accurately?
Undersampling Artifact
MRI GMAIMMbench Organ Recognition - Abdomen
L0 (Original)
L1 (Moderate)
L2 (Severe)
This is a MRI image. Which of the following options is the most appropriate to describe the marked area?
Motion Interference
Object rotation and object movement
Object Movement
MRI OmniMedVQA Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
What is the visual finding in this image?
Object Rotation
X-ray GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Given the boxed region in the X-ray image, which abnormality is the image most indicative of?
Intensity Jitter
Brightness adjustment, exposure, and contrast reduction
Adjust Brightness
Endoscopy GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Given the boxed region in the endoscopy image, which abnormality is the image most indicative of?
Exposure
X-ray GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Focus on the square-highlighted area of this X-ray image. What could be the potential diagnosis?
Reduce Contrast
Ultrasound GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Given that this is a UltraSound image, which option would be the most fitting for the marked area?
Noise
Gaussian noise and low dose artifacts
Gaussian Noise
X-ray GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Focus on the square-highlighted area of this X-ray image. What could be the potential diagnosis?
Low Dose
MRI GMAIMMbench Organ Recognition - Abdomen
L0 (Original)
L1 (Moderate)
L2 (Severe)
Observe the MRI image. Can you identify the organ in the highlight area?
Resolution & Blur
Low resolution, motion blur, Gaussian blur, bubble, and X-ray motion blur
Bubble
Histopathology GMAIMMbench Cell Recognition
L0 (Original)
L1 (Moderate)
L2 (Severe)
Which cell type is indicated in the highlighted area of the Histopathology image?
Low Resolution
X-ray GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
In the area enclosed by a box in this X-ray image, what pathology is most likely present?
Gaussian Blur
Endoscopy GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Focus on the square-highlighted area of this endoscopy image. What could be the potential diagnosis?
Motion Blur
Endoscopy GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Observe the endoscopy image focusing on the area within the box. What is the most likely abnormality depicted in this section?
X-ray Motion Blur
X-ray GMAIMMbench Disease Diagnosis
L0 (Original)
L1 (Moderate)
L2 (Severe)
Considering the box-marked region in the X-ray image, what is the most likely diagnosis?
Want to Try on Your Own Data?
Download the full benchmark dataset with 24,894 QA pairs across 18 degradation types.