Irina Grigorescu

Artificial Intelligence
How deep learning can help predict cancer recurrence
Researchers have shown that a deep-learning algorithm can be used to extract interpretable features from annotation-free histopathology images from prostate cancer patients. Their framework outperformed the prediction of biochemical recurrence using conventional methods.
February 13, 2020
2020 02 13 19 51 3108 2020 02 14 Pathology 20200213191628
Page 1 of 1
Top Stories
Cardiac MRI analysis can provide useful information on epicardial adipose tissue, infarct size, microvascular obstruction, and myocardial edema.
Cardiovascular Radiology
MRI casts light on patient outcome after heart attack
Increased volume of fat tissue around the heart tends can be associated with greater myocardial damage after a heart attack, German researchers have reported.
A smartphone is used to take an image for a 3D-printed prosthesis.
Advanced Visualization
3D-printed prostheses bring hope in conflict zones
Obesity Text2
MRI
MRI proves effective for tracking results of obesity interventions
Croatia Brushstroke Flag
CT
Lung cancer screening: How Croatia does it
Example case from test set with biparametric MRI data, reference segmentation, and output from AI system. The primary MRI assessment resulted in a PI-RADS 5 lesion in the left dorsal part of the peripheral zone and a PI-RADS 4 lesion in the right dorsal part of the peripheral zone (PSA of 11.6 μg/L). Targeted biopsies from both areas showed ISUP-grade 3. All images are from the same slice location, showing a part of both lesions. A T2-weighted (T2W) image. B Diffusion-weighted image with a b-value of 1,500 s/mm2. (C) Apparent diffusion coefficient map. (D) Reference segmentation with white areas representing tumor. (E) Softmax output from AI system. Inset in the lower left corner shows the detection map (cropped around the prostate) with red voxels representing softmax values of 0.99 and yellow voxels representing softmax values of 0.59. (F) T2W image with reference (D) and detection map (inset in E) overlayed, green represents areas in which both the reference and the AI output are positive (true positive voxels for AI), red only positive in reference (false-negative voxels for AI) and blue only segmented by AI (false-positive voxels for AI).
MRI
AI comes under scrutiny in Swedish prostate MRI study
More in Home
MRI casts light on patient outcome after heart attack
By Maryam Payne
Increased volume of fat tissue around the heart tends can be associated with greater myocardial damage after a heart attack, German researchers have reported.
December 12, 2025
Cardiac MRI analysis can provide useful information on epicardial adipose tissue, infarct size, microvascular obstruction, and myocardial edema.
3D-printed prostheses bring hope in conflict zones
By Philip Ward
3D-printing allows for faster, lighter, more affordable, and highly customized prosthetic production, and the technology can help patients regain mobility.
December 11, 2025
A smartphone is used to take an image for a 3D-printed prosthesis.
MRI proves effective for tracking results of obesity interventions
By Kate Madden Yee
MRI is a powerful tool to monitor skeletal muscle adaptations during lifestyle interventions for obesity, researchers noted.
December 10, 2025
Obesity Text2
Lung cancer screening: How Croatia does it
By Maryam Payne
The Croatian lung cancer screening program, integrated into the national system and designed for equitable access, could serve as a model for other countries.
December 10, 2025
Croatia Brushstroke Flag
AI comes under scrutiny in Swedish prostate MRI study
By Philip Ward
An AI system trained in a screening population has lower specificity than radiologists using PI-RADS v2, but considerable care is vital, researchers from Gothenburg reported in European Radiology on 8 December.
December 9, 2025
Example case from test set with biparametric MRI data, reference segmentation, and output from AI system. The primary MRI assessment resulted in a PI-RADS 5 lesion in the left dorsal part of the peripheral zone and a PI-RADS 4 lesion in the right dorsal part of the peripheral zone (PSA of 11.6 μg/L). Targeted biopsies from both areas showed ISUP-grade 3. All images are from the same slice location, showing a part of both lesions. A T2-weighted (T2W) image. B Diffusion-weighted image with a b-value of 1,500 s/mm2. (C) Apparent diffusion coefficient map. (D) Reference segmentation with white areas representing tumor. (E) Softmax output from AI system. Inset in the lower left corner shows the detection map (cropped around the prostate) with red voxels representing softmax values of 0.99 and yellow voxels representing softmax values of 0.59. (F) T2W image with reference (D) and detection map (inset in E) overlayed, green represents areas in which both the reference and the AI output are positive (true positive voxels for AI), red only positive in reference (false-negative voxels for AI) and blue only segmented by AI (false-positive voxels for AI).
Molecular imaging can trigger 'next revolution' in radiology
By Amerigo Allegretto
Molecular imaging will play a key role in precision medicine, and radiology will be at the forefront in ensuring the best patient outcomes, predicts RSNA President Dr. Umar Mahmood, PhD.
December 8, 2025
Screenshot 82 n L Dg Mpjn V1 Thumbnail
Challenging inequity: ESR group spells out why it matters
By Maryam Payne
Radiology staff must recognize how inequities in imaging affect patients from minority groups, says the ESR subcommittee on equity, diversity, and inclusivity.
December 8, 2025
Color Profiles Res
PET/MRI detects ‘unknown’ generators of musculoskeletal pain
By Will Morton
Potential pain generators were identified by increased F-18 FDG radiotracer uptake on PET and/or structural abnormalities on MRI.
December 5, 2025
Back Pain 400
Digital workflows, 3D models enhance preoperative planning
By Maryam Payne
Low-cost digital workflows using 3D printing and AI-assisted model-building can prove valuable for preoperative planning.
December 5, 2025
Postoperative 3D volume rendering of the fenestrated endograft reconstruction: 3D volume-rendered reconstruction obtained in Horos using the standard soft-tissue preset. The image displays the postoperative appearance of the fenestrated endograft and visceral stents, extending from the descending thoracic aorta to the iliac bifurcation, acquired during the follow-up CT angiogram one month after the procedure. The reconstruction shows patent visceral branches with well-perfused stents, absence of endoleak, and successful restoration of the previously disconnected left limb with adequate iliac perfusion.
Cryoablation treats early-stage breast cancer in older women
By Amerigo Allegretto
Cryoablation is effective in treating early-stage breast cancer in older women who do not undergo surgery, a team from Madrid reported at RSNA 2025.
December 4, 2025
Rsna 2025 Welcome
Stent procedure proves safe in pediatric liver transplant cases
By Will Morton
Biodegradable stents appear safe and effective for treating narrowing of the bile ducts in children after liver transplants, Italian researchers noted at RSNA 2025.
December 4, 2025
Biliary Atresia Disease Adobe Pikovit
Radiology must persevere with sustainability efforts
By Kate Madden Yee
It's crucial to reduce medical imaging's energy use and waste to mitigate the effect of global warming, warns RSNA 2025 keynote speaker Dr. Geraldine McGinty.
December 3, 2025
Sustainability Image
Page 1 of 66
Next Page