AJR Podcasts

Full article: https://www.ajronline.org/doi/10.2214/AJR.23.29670 Sid Dogra, MD discusses an AJR article in which the authors show that retained biopsy site markers following resection of breast lesions increases risk of residual malignancy, particularly in patients whose initial pathology suggested malignancy. Radiologists can improve patient care by more explicitly noting missing biopsy site markers and recommending early postoperative mammogram on initial intraoperative specimen radiographs.

Full article: https://www.ajronline.org/doi/10.2214/AJR.23.29655 Shruti Kumar, MD discusses an AJR study that compares the performance of mammography alone, standalone artificial intelligence (AI), ultrasound alone, and a combination of these methods in a sample of 1,325 women. The study suggests that adding AI to mammography and ultrasound does not provide significant benefit in cancer detection rates in patients with dense breasts. However, the study's limitations, including single-institution data and specific AI platform, must be considered before generalizing the results.

Full article: https://www.ajronline.org/doi/10.2214/ajr.23.29570 Childhood obesity is a growing epidemic globally. Precisely quantifying abdominal fat distribution on MRI could be valuable for research and clinical care, but manual segmentation is extremely tedious. Farzaneh Ghazi Sherbaf, MD discusses a recent study in which the authors developed and validated an AI method to automatically segment subcutaneous and visceral abdominal fat on MRI in adolescents. The work has implications for making obesity research and care more effective through automated, personalized abdominal fat mapping.

Explore a study using deep learning to estimate liver fat fraction in an American Indian population. Compare the accuracy of the two-point Dixon method with a CNN derived method in assessing liver iron overload. Discuss a study proposing a method to quantify liver fat deposition in line with standard abdominal imaging protocols.

Full article: https://www.ajronline.org/doi/10.2214/ajr.23.29674 Peter Gunderman, MD discusses a recent study that finds artificial intelligence can accurately predict the pathology of pure ground glass nodules, at least in the subset of nodules representing variants of adenocarcinoma.

Full article: https://www.ajronline.org/doi/abs/10.2214/AJR.23.29609 Hong Linda Li, MD discusses a new AJR article that details how preoperative MRI-based staging can predict biochemical recurrence of prostate cancer after radical prostatectomy as effectively as post-surgical pathologic staging.

Shambo Guha Roy, MD discusses a deep learning model for predicting mortality in community-acquired pneumonia based on chest radiographs. The model outperforms traditional scoring systems, offering potential for improved risk assessment in clinical settings.

Full article: https://www.ajronline.org/doi/10.2214/AJR.23.29340 Wenhui Zhou, MD, PhD reviews a recent AJR article that explored the feasibility of pulmonary MRA for evaluating pulmonary embolism during a contrast media shortage in 2022. The results of this study highlight the value of pulmonary MRA as a practical alternative to pulmonary CTA, especially in emergency settings.

Full article: https://www.ajronline.org/doi/10.2214/AJR.23.29285 Leo Jia, MD discusses a new AJR article that finds radiologic evidence of unambiguous extranodal extension of lymph node metastases may have a role in the initial staging evaluation of patients diagnosed with non-small cell lung cancer.

Full article: https://www.ajronline.org/doi/abs/10.2214/AJR.23.29151 Rachana Borkar, MBBS discusses a study shows high negative predictive value for clear cell renal cell carcinoma. The score could impact clinical decisions in patients by guiding risk stratification and patient selection for active surveillance through identification of patients unlikely to have clear cell renal cell carcinoma.