Image from article in Radiology: Artificial Intelligence
Federated learning, synthetic data generation, de-identification and more! A new review on imaging data preparation for AI https://doi.org/10.1148/ryai.250273 #AI #deID #ML
Image from article in Radiology: Artificial Intelligence
US-derived Pediatric Kidney Length and Volume Percentiles by Age: A Big Data Approach https://doi.org/10.1148/ryai.250056 #NLP #PedsRad #AI
You have to see this #ASPNR26 poster!
Developing alternative teaching methods provides an opportunity to expose preclinical learners to radiology & radiologists and may provide a networking opportunity to encourage medical students to consider a career in radiology.
#ASPNR26 poster: Pediatric parathyroid adenomas are an extremely rare but treatable cause of hyperparathyroidism. There is a bimodal distribution with a smaller early peak in neonates and a larger second peak occurring from 6 years of age through adolescence
#ASPNR26 poster: Inadvertent ingestion of food containing grill bristle foreign bodies causes acute and severe symptoms of sore throat, odynophagia, and/or globus sensation.
Vertebral level is a more accurate landmark than the kidneys for assessing conus position on fetal MRI.
Make sure you read this #ASPNR26 poster from Dr. Tracie Kong and the @CincyKidsRad #FetalImaging team to learn more.
¡Falta 1 semana! No se pierda el curso #MSK en Perú (23-24 de enero). Casos prácticos, aprendizaje interactivo y Q&A con expertos internacionales. #RSNASpotlight https://bit.ly/43j98Gp @andrewtrout.bsky.social
Slide shows combinations of AI and radiologist interpretation - whenever there is doubt by one the final outcome is doubt
Saurabh Jha at #RSNA2025
The problem of Radiologist + AI:
My first time at #slarp and my first time in #bolivia - enjoying meeting #pedsrad from across South America & relearning #pedsrad in Spanish
Are we unnecessarily withholding IV contrast from patients with kidney disease?
- Risk of contrast-induced nephropathy may be overstated
- Withholding contrast can delay diagnosis
This may be a #TWDFNR!
🔗 doi.org/10.1002/jhm.70063
🖼️ Visual Abstract by C Coleman + JHM Team
So this is kinda cool…
We know #AI built for adults struggles in #pediatrics
Metadata spoofing (tricking the algo to think its seeing an adult size patient) improves segmentation performance
Simple but effective - kudos to Jon Dudley for thinking of this
#radsky
www.ajronline.org/doi/abs/10.2...
A novel clavicle fracture dating atlas significantly improved radiologists' performance and agreement in dating not only clavicle fractures but also humeral and femoral fractures in infants, offering a valuable tool in suspected child abuse cases.
www.ajronline.org/doi/10.2214/...
A new generation of radiotherapies promises a more targeted attack on cancer.
www.science.org/content/arti...
This multi-institutional study finds that ultrasound has excellent diagnostic performance for midgut malrotation and volvulus in children, supporting its use as a first-line imaging modality for these conditions.
www.ajronline.org/doi/10.2214/...
Image from article in Radiology: Artificial Intelligence
#CoRaX demonstrated potential to aid radiologists in detecting abnormalities on CXRs by identifying & correcting perceptual errors https://doi.org/10.1148/ryai.240277 @uh.edu #CIRSE2025 #IRad #CVRad #radiology
The GOP said its overhaul of Medicaid was aimed at reducing fraud and getting more adult beneficiaries to work. Among the likely side effects: fewer services and doctors for treating sick children.
Image from article in Radiology: Artificial Intelligence
CLAIM: the Checklist for AI in Medical Imaging
#guideline #checklist https://rsna.org/claim #MachineLearning #DeepLearning #Radiomics
Donate it to their local @npr.org station!
Academic radiologists are increasingly withdrawing discretionary effort—engaging only in their baseline duties—a phenomenon known as "quiet quitting".
www.ajronline.org/doi/10.2214/...
I describe the structure of a manuscript to trainees as having the shape of the figure on the right. Intro & Discussion provide broad context while the bulk of the rest of the paper is at the waist of the hourglass (credit to my uncle the biologist). “Presentation” verbal or written has many forms.
A first-rate review of MASH (fatty liver disease) @nejm.org www.nejm.org/doi/full/10....
Both #EUS & #MRI are acceptable imaging modalities for diagnosis of chronic #pancreatitis in #children - we wanted to understand how findings on these modalities might match up (or not)
=the modalities agree on duct findings but less so on parenchymal findings
www.sciencedirect.com/science/arti...
Tips for asking ?s as moderator #ISTH2025
Clinical talk: what future directions do you see this work taking?
Basic science: how do you envision this translating to the clinic?
Next level pro tip: contact author beforehand to see what they didn’t have time to present that they’d like you to ask
If we want a stronger Ohio tomorrow, we have to invest in our children today. That starts with supporting our teachers, investing in our schools, and giving every child a chance to succeed.
NIH staff write urgent open letter to the NIH Director:
"We urge you as NIH Director to restore grants delayed or terminated for political reasons so that life-saving science can continue."
www.standupforscience.net/bethesda-dec...
Absolutely! Editors need to make faster decisions (even if desk reject) and not let authors wait for months
Dear editors, your publication delays are damaging our careers
www.nature.com/articles/d41...
Learn how we are battling #Cancer at Cincinnati Children’s with innovations to accelerate treatments for #Leukemia, develop #CancerImmuneTherapies, and more. #RAR24 #MedSky #PedsSky #SciSky Check out our latest Research Annual Report: scienceblog.cincinnatichildrens.org/research-exc...
I missed DDW…
Looks like more interventions in patients with SCD.
Were attacks of pancreatitis more severe?
I wonder about the impact of iron loading on attack severity.
Things We Do For No Reason: Avoiding IV iodinated contrast in AKI or CKD
I think of those who give prolonged meropenem for "undifferentiated sepsis" but refuse to arrange contrast CT to look for the source because the patient has abnormal RFT
doi.org/10.1002/jhm....
#IDSky
