
How to Fix a Blurry Ultrasound Photo: Enhance Baby Scan Images for Keepsakes
Ultrasound photos look blurry and grainy by nature β but some of that degradation is fixable. Learn what AI can improve (JPEG artifacts, screen photography noise, low resolution) and what it can't (inherent speckle), plus how to get a better copy from your hospital.
Dr. Anna Chen
Quick path: If your ultrasound photo is a blurry phone shot of the screen or a heavily-compressed JPEG from the patient portal, the secondary degradation is fixable. Try the JPEG artifact remover, photo denoiser, or photo enhancer β $4.99 one-time, no subscription.
An ultrasound photo is one of the first images you will ever have of your child. For many families it becomes a lasting keepsake β printed, framed, and shared with grandparents before anyone has met the baby. Which makes it especially frustrating that most ultrasound photos look technically poor: dim, grainy, soft, and riddled with compression artifacts.
Understanding why they look this way β and which problems are actually fixable β requires a brief detour into how sonogram technology works and the different stages at which image quality degrades on the way from the machine to your hands.
Why Do Ultrasound Photos Look So Blurry and Grainy?
The fundamental cause is that ultrasound imaging uses sound, not light. A conventional camera captures photons reflecting off surfaces; the sonogram machine emits high-frequency sound pulses into tissue, measures the time and intensity of returning echoes, and reconstructs a cross-sectional image from acoustic data. This process has a hard resolution ceiling set by physics: acoustic wavelength, scan line density, tissue depth, and echo attenuation.
The characteristic grainy texture in any sonogram β technically called speckle β is not damage or noise in the conventional sense. It is an inherent property of the acoustic interference pattern in the returning echoes. Speckle is part of the image content, and it carries diagnostic information that radiologists and sonographers read.
However, by the time a patient receives their copy of the scan, several layers of additional degradation have typically accumulated on top of this inherent speckle:
Thermal print fading: Most ultrasound machines print to thermal paper β the same technology as a store receipt. Thermal prints begin fading immediately and can become difficult to read within a few years, even when stored away from light.
Hospital JPEG compression: When hospitals export ultrasound images through patient portals or email, they typically save as JPEG at moderate quality settings. JPEG compression is lossy and introduces block-pattern artifacts, particularly in the low-contrast gradient regions common in sonograms.
Phone photography of the screen: Many patients simply photograph the ultrasound monitor with their smartphone during the appointment. This adds the phone's own sensor noise, blur from hand movement or screen curvature, and potential moirΓ© patterns from the screen's pixel grid interacting with the camera sensor.
Each of these is a separate degradation layered on top of the original image.
What Can AI Enhancement Realistically Improve?
The secondary degradations β JPEG artifacts, screen photography noise, low resolution from scanned prints β are all addressable. The inherent speckle is not, and attempts to remove it aggressively would blur the diagnostic content of the image.
JPEG Artifacts from Hospital Email Export
If your ultrasound image came from a patient portal or email and shows a blocky, pixelated quality especially in the darker gradient regions, it has JPEG compression artifacts. The JPEG artifact remover uses SwinIR, a transformer model specifically trained on compression artifact removal, to smooth the block-pattern distortion without blurring real edge structure. On a JPEG-compressed sonogram, this typically recovers the cleaner gradient transitions visible in the original before compression.
Noise and Blur from Screen Photography
A smartphone photo of an ultrasound monitor has multiple degradation sources: sensor noise (especially in the dim room lighting of a sonogram suite), blur from hand movement or screen curvature, and color cast from room lighting. The photo denoiser targets the sensor noise layer using NAFNet, which can distinguish real edge structure from noise. The photo deblurrer recovers edge sharpness from mild motion blur. Together they can recover a substantially cleaner version of a phone-shot screen photo.
Low Resolution from Scanned Thermal Prints
If you have scanned a thermal print (or photographed one), the source resolution may be quite low relative to what you want to display or print. The photo enhancer upscales via Real-ESRGAN β a model trained for natural image super-resolution β which can increase pixel dimensions 4Γ while adding plausible high-frequency detail. The result prints and displays more cleanly than a simple bicubic upscale.
What Can AI Not Fix in an Ultrasound Photo?
It is important to set realistic expectations before enhancing a sonogram keepsake.
Inherent speckle: The grainy texture that defines the look of an ultrasound image is not removable without degrading the image content itself. The photo denoiser is calibrated for photographic sensor noise, which has different statistical properties from acoustic speckle. Standard processing will not over-remove speckle, but very aggressive denoising settings would begin to blur it. Conservative settings are appropriate.
Very soft or unrecognizable features: If your ultrasound shows the baby in a position where facial features are partially obscured by tissue depth or scanning angle, AI enhancement cannot reconstruct features that were not acoustically captured. Enhancement recovers image quality; it does not reconstruct missing content.
Extreme fade on old thermal prints: A thermal print that has faded to very low contrast β where the image is barely distinguishable from the background β has lost the image information itself. Contrast enhancement can help recover what remains, but there is an information floor below which nothing can be recovered. The old photo restoration tool handles this better than the standard enhancer for severely faded prints.
What Is the Best Step-by-Step Process for Enhancing an Ultrasound Keepsake?
Starting point A β Phone photo of the screen:
- Photo denoiser β remove sensor noise and room lighting cast
- Photo deblurrer β recover edge sharpness if soft
- Photo enhancer β upscale to full keepsake resolution
Starting point B β JPEG from hospital portal:
- JPEG artifact remover β remove block-pattern compression artifacts
- Photo enhancer β upscale for printing
Starting point C β Scanned or photographed thermal print:
- Old photo restoration β if contrast is very low
- Photo denoiser β clean up scan noise
- Photo enhancer β upscale
In all cases, use conservative settings. A clean version of the real image is a better keepsake than a smooth version that no longer looks like a sonogram.
How Do You Get a Better Digital Copy from the Hospital?
The most effective improvement happens before any software is involved. Ask your hospital imaging department for:
-
A direct DICOM export: Most modern ultrasound machines save to DICOM, a lossless medical format. DICOM files contain the full-resolution sonogram without JPEG compression. Patients are generally entitled to request their own imaging records in any format the hospital system can produce.
-
A high-quality PNG or TIFF: If DICOM is not available, ask for the highest-quality lossless export the system supports. Even a large JPEG at quality 95+ is substantially better than a portal export at quality 70.
-
A print from the imaging workstation: Some departments can print directly from the workstation at higher quality than the standard thermal print roll.
A better source file produces dramatically better enhancement results. The photo enhancer can recover significant quality from a degraded source, but it cannot add information that was lost in compression before you received the file.
Related reading:
- JPEG Artifact Remover β remove compression block patterns
- Photo Denoiser β clean up screen photography noise
- Photo Deblurrer β sharpen soft photos
- Photo Enhancer β upscale and enhance, $4.99 one-time
- Old Photo Restoration β for faded thermal prints
- Photo Colorizer β add color to black-and-white keepsake photos
About the Author
Dr. Anna Chen
Maternal Health Educator & Medical Imaging Consultant
Dr. Anna Chen holds advanced training in maternal health education and medical imaging systems. She has worked with expectant families for over twelve years, helping them understand prenatal imaging technology and preserve their pregnancy keepsakes with accuracy and care.
Share this article
Ready to Restore Your Old Photos?
Try ArtImageHub's AI-powered photo restoration. Bring faded, damaged family photos back to life in seconds.