Imaging Informatics

CT-based AI model predicts pneumonitis in lung cancer patients

By Amerigo Allegretto

The algorithm outperformed clinical and radiomics models in these patients.

September 25, 2026

CIPHER, an AI-powered CT foundation model, identifies patients with lung cancer at increased risk of immunotherapy-induced pneumonitis before treatment begins. The AI-generated output, shown here, highlights lung abnormalities to support early risk assessment and personalized patient care.

PACS workstations add to power consumption in radiology departments

By Amerigo Allegretto

PACS workstations consume about 25 kW when active while also consuming power cumulatively in powered down states.

September 23, 2026

Radiologist Dictating (1)

MRI-based model shows changes over time in multiple myeloma patients

By Amerigo Allegretto

Whole-body MRI could serve as a scalable biomarker for risk stratification in multiple myeloma patients.

September 4, 2026

Multiple Myeloma 400

eConsult systems may benefit radiology-provider communication

By Amerigo Allegretto

Electronic consult (eConsult) platforms provide documented pathway for imaging questions and support radiologists as consultants.

September 3, 2026

Doctor Patient Consult Tablet 400

AI cuts chest CT reporting time

Exploratory modeling suggested an approximate reduction of 0.5 full-time equivalent radiologist workload.

September 1, 2026

Example of AI-assisted pulmonary nodule detection and longitudinal volumetric assessment in a 65-year-old woman with peripheral arterial disease, a 24-pack-year smoking history, emphysema, and a family history of lung cancer. An incidental 8.5-mm right lower lobe pulmonary nodule is identified at CT angiography of the abdomen in July 2022 but does not receive follow-up. Subsequent imaging demonstrates interval growth, and CT-guided biopsy confirms stage IA lung adenocarcinoma. Contrast-enhanced axial chest CT images obtained on February 2, 2024, before robotic-assisted thoracoscopic segmentectomy, show a 12.1-mm spiculated right lower lobe pulmonary nodule. (A) Conventional contrast-enhanced axial chest CT image without AI assistance shows a 12.1-mm spiculated right lower lobe pulmonary nodule. (B) Corresponding CT image with AI-assisted pulmonary nodule detection, automated segmentation, and quantitative nodule assessment identifies a solid pulmonary nodule (nodule #1) measuring 13 × 8 (11) mm, with a volume of 739 mm3, a 45% volumetric increase compared with the previous contrast-enhanced chest CT performed on October 9, 2023 (510 mm3), and an estimated volume doubling time (VDT) of 216 days. (C) AI-generated longitudinal nodule analysis displays automated segmentation of the current (February 2, 2024) and previous (October 9, 2023) CT examinations, 3D nodule reconstruction, and quantitative volumetric assessment for longitudinal follow-up. The AI interface is displayed in the original software language (Dutch).

Most FDA-cleared AI medical devices not tested on patient outcomes

By Amerigo Allegretto

Of the 1,059 FDA-cleared devices used in radiology, three had a registered trial.

August 19, 2026

Artificial Intelligence Ai Global Communication Network 400

AI framework could improve MRI quality control in prostate cancer

By Amerigo Allegretto

The framework uses a DL model trained on low- and high-quality MR images and helped improve radiologist accuracy.

August 19, 2026

Low-quality axial T2-weighted MRI scans with gradient-weighted class activation mapping. The low-quality MRI scans (top row) show blurred structures and artifacts making zonal and capsule boundaries indistinguishable. On the same scans with gradient-weighted class activation map overlays (bottom row), red indicates regions that were most heavily weighted by the deep learning model toward the low-quality prediction.

Longitudinal DBT could help improve long-term cancer risk prediction

By Amerigo Allegretto

A longitudinal DBT model outperformed other models based on mammography and clinical factors in predicting five-year cancer risk.

August 13, 2026

AI tools may not meet breast radiologists' expectations in practice

By Amerigo Allegretto

AI tools in breast imaging may have more modest clinical impact on recall and biopsy rates.

August 12, 2026

Breast Mammo Computer 400 (3)

Model confidence, reader expertise affect reader-LLM collaboration

By Amerigo Allegretto

Successful human-LLM collaboration is associated with model confidence and reader expertise in chest imaging.

August 11, 2026

Example cases illustrate the double-edged role of large language model (LLM) rationale quality. (A) A 50-year-old female patient with a palpable anterior chest wall mass (correct diagnosis: solitary fibrous tumor of the pleura). The LLM incorrectly suggested synovial sarcoma but provided a persuasive rationale (quality score: 4 of 5), describing a well-defined nodule at axial contrast-enhanced CT spanning the pleura and intercostal space with heterogeneous strong enhancement. Four of five readers (80%) accepted the incorrect suggestion. FDG = fluorodeoxyglucose. (B) A 61-year-old male patient with diabetes presented with cough, sputum, and dyspnea (correct diagnosis: mucormycosis). The LLM correctly suggested mucormycosis with an excellent rationale (quality score: 5 of 5), integrating the reverse halo sign on serial axial CT images and rapidly progressive cavitary lesions. Three of five readers initially selected incorrect diagnoses; after LLM output, all five selected the correct diagnosis (100% acceptance). Blue and orange reader icons denote high- and low-expertise readers, respectively.

Many women interested in AI-based risk assessment for breast cancer

By Amerigo Allegretto

Survey findings also show that most women at higher risk are interested in additional screening tests.

August 10, 2026

Breast Cancer 400

DBT-based foundational model classifies breast density

By Amerigo Allegretto

DBT-DINO showed strong performance in classifying breast density from DBT images.

August 4, 2026

(A) Free-response receiver operating characteristic (FROC) curve for DBT-DINO and DINOv2 (Meta AI) on the detection task. DINOv2 achieved higher sensitivity than DBT-DINO across most false-positive thresholds. (B) Detection rates of DBT-DINO and DINOv2 for lesions that were ultimately found to be benign (n = 70) and lesions that were found to be cancerous (n = 66). Error bars represent 95% CIs estimated via patient-level bootstrapping with 1000 repetitions. Both models showed similar detection rates for cancerous lesions, whereas DINOv2 detected more benign lesions than DBT-DINO. (C) Example bounding box predictions for DBT-DINO and DINOv2 on the detection test set. Each image is a single two-dimensional section from a three-dimensional digital breast tomosynthesis volume; the first column shows a mediolateral oblique view, and the second and third columns show craniocaudal views. Both models struggled to detect very small lesions, such as the 3-mm-diameter lesion shown in the first column. FP = false positive.

Page 1 of 1858