Bidirectional Encoder Representations from Transformers in Radiology: A Systematic Review of Natural Language Processing Applications.

INTRODUCTION: Bidirectional Encoder Representations from Transformers (BERT), introduced in 2018, has revolutionized natural language processing. Its bidirectional understanding of word context has enabled innovative applications, notably in radiology. This study aimed to assess BERT's influence and applications within the radiologic domain. METHODS: Adhering to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, we conducted a systematic review, searching PubMed for literature on BERT-based models and natural language processing in radiology from January 1, 2018, to February 12, 2023. The search encompassed keywords related to generative models, transformer architecture, and various imaging techniques. RESULTS: Of 597 results, 30 met our inclusion criteria. The remaining were unrelated to radiology or did not use BERT-based models. The included studies were retrospective, with 14 published in 2022. The primary focus was on classification and information extraction from radiology reports, with x-rays as the prevalent imaging modality. Specific investigations included automatic CT protocol assignment and deep learning applications in chest x-ray interpretation. CONCLUSION: This review underscores the primary application of BERT in radiology for report classification. It also reveals emerging BERT applications for protocol assignment and report generation. As BERT technology advances, we foresee further innovative applications. Its implementation in radiology holds potential for enhancing diagnostic precision, expediting report generation, and optimizing patient care.

Predictive factors for radiation-induced pituitary damage in pediatric patients with brain tumors.

BACKGROUND AND PURPOSE: Multiple studies demonstrated hypothalamic-pituitary dysfunction in survivors of pediatric brain tumors. However, few studies investigated the trajectories of pituitary height in these patients and their associations with pituitary function. We aimed to evaluate longitudinal changes of pituitary height in children and adolescents with brain tumors, and their association with endocrine deficiencies. MATERIALS AND METHODS: We conducted a retrospective analysis of 193 pediatric patients (54.9% male) diagnosed with brain tumors from 2002 to 2018, with a minimum of two years of radiological follow-up. Pituitary height was measured using MRI scans at diagnosis and at 2, 5, and 10 years post-diagnosis, with clinical data sourced from patient charts. RESULTS: Average age at diagnosis was 7.6 ± 4.5 years, with a follow-up of 6.1 ± 3.4 years. 52.8% underwent radiotherapy and 37.8% experienced pituitary hormone deficiency. Radiation treatment was a significant predictor of decreased pituitary height at all observed time points (p = 0.016, p < 0.001, p = 0.008, respectively). Additionally, chemotherapy (p = 0.004) or radiotherapy (p = 0.022) history and pituitary height at 10 years (p = 0.047) were predictors of endocrine deficiencies. ANOVA revealed an expected increase in pituitary height over time in pediatric patients, but this growth was significantly impacted by radiation treatment and gender (p for interaction = 0.005 and 0.025, respectively). CONCLUSION: Cranial irradiation in pediatric patients is associated with impairment of the physiologic increase in pituitary size; in turn, decreased pituitary height is associated with endocrine dysfunction. We suggest that pituitary gland should be evaluated on surveillance imaging of pediatric brain tumor survivors, and if small for age, clinical endocrine evaluation should be pursued.

A Novel Artificial Intelligence Based Denoising Method for Ultra-Low Dose CT Used for Lung Cancer Screening.

RATIONALE AND OBJECTIVES: To assess ultra-low-dose (ULD) computed tomography as well as a novel artificial intelligence-based reconstruction denoising method for ULD (dULD) in screening for lung cancer. MATERIALS AND METHODS: This prospective study included 123 patients, 84 (70.6%) men, mean age 62.6 ± 5.35 (55-75), who had a low dose and an ULD scan. A fully convolutional-network, trained using a unique perceptual loss was used for denoising. The network used for the extraction of the perceptual features was trained in an unsupervised manner on the data itself by denoising stacked auto-encoders. The perceptual features were a combination of feature maps taken from different layers of the network, instead of using a single layer for training. Two readers independently reviewed all sets of images. RESULTS: ULD decreased average radiation-dose by 76% (48%-85%). When comparing negative and actionable Lung-RADS categories, there was no difference between dULD and LD (p = 0.22 RE, p > 0.999 RR) nor between ULD and LD scans (p = 0.75 RE, p > 0.999 RR). ULD negative likelihood ratio (LR) for the readers was 0.033-0.097. dULD performed better with a negative LR of 0.021-0.051. Coronary artery calcifications (CAC) were documented on the dULD scan in 88(74%) and 81(68%) patients, and on the ULD in 74(62.2%) and 77(64.7%) patients. The dULD demonstrated high sensitivity, 93.9%-97.6%, with an accuracy of 91.7%. An almost perfect agreement between readers was noted for CAC scores: for LD (ICC = 0.924), dULD (ICC = 0.903), and for ULD (ICC = 0.817) scans. CONCLUSION: A novel AI-based denoising method allows a substantial decrease in radiation dose, without misinterpretation of actionable pulmonary nodules or life-threatening findings such as aortic aneurysms.

Quantitative and qualitative analysis of fetal temporal lobe T2 signal in cytomegalovirus infected fetuses and normal controls.

BACKGROUND: Temporal lobe T2 hyperintensity has been described in association with prenatal cytomegalovirus (CMV) infection on fetal MRI. However, these findings are often perplexing with high inter-observer variability. Our objective was to evaluate temporal lobe T2 signal quantitatively in prenatal CMV infection. METHODS: In this retrospective study, 119 fetuses, of which 51 fetuses with suspected CMV exposure (29-36 weeks of gestation) based on maternal seroconversion and age-matched 68 normal controls, were included. Mean and maximal temporal lobe T2 signal were evaluated quantitatively by measuring the T2 signal in the temporal lobes relative to the amniotic fluid's signal. Intra-observer, inter-observer variability and diagnostic performance were assessed. The occurrence of neonatal sensorineural hearing loss (SNHL) was recorded. RESULTS: Relative temporal lobe T2 signal did not change along with the examined gestational age. Of our suspected CMV cohort, 29 fetuses were positive for fetal CMV infection on polymerase chain reaction (PCR) analysis. There were no statistically significant differences in the relative mean or maximal temporal lobes T2 signal between CMV positive, CMV negative fetuses, or normal controls. No correlation was found between neonatal SNHL and temporal lobe T2 signal. CONCLUSIONS: When temporal lobe T2 signal is analyzed quantitatively, CMV infected fetuses do not present an increased signal than age-matched controls. Thus, reported subjective temporal T2 hyperintensities should be interpreted carefully and should have a limited effect on pregnancy management, especially as an isolated finding. Our study illustrates the importance of quantitative imaging in diagnostic neuroradiology.