Artificial intelligence for X-ray scaphoid fracture detection: a systematic review and diagnostic test accuracy meta-analysis.

OBJECTIVES: Scaphoid fractures are usually diagnosed using X-rays, a low-sensitivity modality. Artificial intelligence (AI) using Convolutional Neural Networks (CNNs) has been explored for diagnosing scaphoid fractures in X-rays. The aim of this systematic review and meta-analysis is to evaluate the use of AI for detecting scaphoid fractures on X-rays and analyze its accuracy and usefulness. MATERIALS AND METHODS: This study followed the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) and PRISMA-Diagnostic Test Accuracy. A literature search was conducted in the PubMed database for original articles published until July 2023. The risk of bias and applicability were evaluated using the QUADAS-2 tool. A bivariate diagnostic random-effects meta-analysis was conducted, and the results were analyzed using the Summary Receiver Operating Characteristic (SROC) curve. RESULTS: Ten studies met the inclusion criteria and were all retrospective. The AI's diagnostic performance for detecting scaphoid fractures ranged from AUC 0.77 to 0.96. Seven studies were included in the meta-analysis, with a total of 3373 images. The meta-analysis pooled sensitivity and specificity were 0.80 and 0.89, respectively. The meta-analysis overall AUC was 0.88. The QUADAS-2 tool found high risk of bias and concerns about applicability in 9 out of 10 studies. CONCLUSIONS: The current results of AI's diagnostic performance for detecting scaphoid fractures in X-rays show promise. The results show high overall sensitivity and specificity and a high SROC result. Further research is needed to compare AI's diagnostic performance to human diagnostic performance in a clinical setting. CLINICAL RELEVANCE STATEMENT: Scaphoid fractures are prone to be missed secondary to assessment with a low sensitivity modality and a high occult fracture rate. AI systems can be beneficial for clinicians and radiologists to facilitate early diagnosis, and avoid missed injuries. KEY POINTS: • Scaphoid fractures are common and some can be easily missed in X-rays. • Artificial intelligence (AI) systems demonstrate high diagnostic performance for the diagnosis of scaphoid fractures in X-rays. • AI systems can be beneficial in diagnosing both obvious and occult scaphoid fractures.

Thymic Imaging Pitfalls and Strategies for Optimized Diagnosis.

Thymic imaging is challenging because the imaging appearance of a variety of benign and malignant thymic conditions are similar. CT is the most commonly used modality for mediastinal imaging, while MRI and fluorine 18 fluorodeoxyglucose (FDG) PET/CT are helpful when they are tailored to the correct indication. Each of these imaging modalities has limitations and technical pitfalls that may lead to an incorrect diagnosis and mismanagement. CT may not be sufficient for the characterization of cystic thymic processes and differentiation between thymic hyperplasia and thymic tumors. MRI can be used to overcome these limitations but is subject to other potential pitfalls such as an equivocal decrease in signal intensity at chemical shift imaging, size limitations, unusual signal intensity for cysts, subtraction artifacts, pseudonodularity on T2-weighted MR images, early imaging misinterpretation, flow and spatial resolution issues hampering assessment of local invasion, and the overlap of apparent diffusion coefficients between malignant and benign thymic entities. FDG PET/CT is not routinely indicated due to some overlap in FDG uptake between thymomas and benign thymic processes. However, it is useful for staging and follow-up of aggressive tumors (eg, thymic carcinoma), particularly for detection of occult metastatic disease. Pitfalls in imaging after treatment of thymic malignancies relate to technical challenges such as postthymectomy sternotomy streak metal artifacts, differentiation of postsurgical thymic bed changes from tumor recurrence, or human error with typical "blind spots" for identification of metastatic disease. Understanding these pitfalls enables appropriate selection of imaging modalities, improves diagnostic accuracy, and guides patient treatment. ©RSNA, 2024 Test Your Knowledge questions for this article are available in the supplemental material.

Large language models for generating medical examinations: systematic review.

BACKGROUND: Writing multiple choice questions (MCQs) for the purpose of medical exams is challenging. It requires extensive medical knowledge, time and effort from medical educators. This systematic review focuses on the application of large language models (LLMs) in generating medical MCQs. METHODS: The authors searched for studies published up to November 2023. Search terms focused on LLMs generated MCQs for medical examinations. Non-English, out of year range and studies not focusing on AI generated multiple-choice questions were excluded. MEDLINE was used as a search database. Risk of bias was evaluated using a tailored QUADAS-2 tool. RESULTS: Overall, eight studies published between April 2023 and October 2023 were included. Six studies used Chat-GPT 3.5, while two employed GPT 4. Five studies showed that LLMs can produce competent questions valid for medical exams. Three studies used LLMs to write medical questions but did not evaluate the validity of the questions. One study conducted a comparative analysis of different models. One other study compared LLM-generated questions with those written by humans. All studies presented faulty questions that were deemed inappropriate for medical exams. Some questions required additional modifications in order to qualify. CONCLUSIONS: LLMs can be used to write MCQs for medical examinations. However, their limitations cannot be ignored. Further study in this field is essential and more conclusive evidence is needed. Until then, LLMs may serve as a supplementary tool for writing medical examinations. 2 studies were at high risk of bias. The study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Advancing Medical Practice with Artificial Intelligence: ChatGPT in Healthcare.

BACKGROUND: Advancements in artificial intelligence (AI) and natural language processing (NLP) have led to the development of language models such as ChatGPT. These models have the potential to transform healthcare and medical research. However, understanding their applications and limitations is essential. OBJECTIVES: To present a view of ChatGPT research and to critically assess ChatGPT's role in medical writing and clinical environments. METHODS: We performed a literature review via the PubMed search engine from 20 November 2022, to 23 April 2023. The search terms included ChatGPT, OpenAI, and large language models. We included studies that focused on ChatGPT, explored its use or implications in medicine, and were original research articles. The selected studies were analyzed considering study design, NLP tasks, main findings, and limitations. RESULTS: Our study included 27 articles that examined ChatGPT's performance in various tasks and medical fields. These studies covered knowledge assessment, writing, and analysis tasks. While ChatGPT was found to be useful in tasks such as generating research ideas, aiding clinical reasoning, and streamlining workflows, limitations were also identified. These limitations included inaccuracies, inconsistencies, fictitious information, and limited knowledge, highlighting the need for further improvements. CONCLUSIONS: The review underscores ChatGPT's potential in various medical applications. Yet, it also points to limitations that require careful human oversight and responsible use to improve patient care, education, and decision-making.

Elevated Native T1 Values in the Remote Myocardium Supplied by Obstructive Non-Infarct Related Coronary Arteries in Post-STEMI Cardiac Magnetic Resonance.

INTRODUCTION: Native T1 mapping values are elevated in acutely injured myocardium. We sought to study whether native T1 values, in the non-infarct related myocardial territories, might differ when supplied by obstructive or nonobstructive coronary arteries. METHODS: Consecutive patients (N = 60, mean age 59 years) with the first STEMI following primary percutaneous coronary intervention, underwent cardiac magnetic resonance within 5 ± 2 days. A retrospective review of coronary angiography reports classified coronary arteries as infarct-related coronary artery (IRA) and non-IRA. Obstructive coronary artery disease (CAD) was defined as stenosis ≥50%. Native T1 values were presented using a 16-segment AHA model according to the three main coronary territories: left anterior descending (LAD), left circumflex (LCX), and right coronary artery (RCA). RESULTS: The cutoff native T1 value for predicting obstructive non-IRA LAD was 1,309 msec with a sensitivity and specificity of 67% and 82%, respectively (AUC 0.76, 95% CI: 0.57-0.95, p = 0.04). The cutoff native T1 value for predicting obstructive non-IRA RCA was 1,302 msec with a sensitivity and specificity of 83% and 55%, respectively (AUC 0.7, 95% CI: 0.52-0.87, p = 0.05). Logistic regression model adjusted for age and infarct size demonstrated that native T1 was an independent predictor for the obstructive non-IRA LAD (OR 4.65; 1.32-26.96, p = 0.05) and RCA (OR 3.70; 1.44-16.35, p = 0.03). CONCLUSION: Elevated native T1 values are independent predictors of obstructive non-IRA in STEMI patients. These results suggest the presence of concomitant remote myocardial impairment in the non-infarct territories with obstructive CAD.

Perivascular Cuffing as the Sole Imaging Finding of Pancreatic Cancer.

BACKGROUND: Perivascular cuffing as the sole imaging manifestation of pancreatic ductal adenocarcinoma (PDAC) is an under-recognized entity. OBJECTIVES: To present this rare finding and differentiate it from retroperitoneal fibrosis and vasculitis. METHODS: Patients with abdominal vasculature cuffing were retrospectively collected (January 2011 to September 2017). We evaluated vessels involved, wall thickness, length of involvement and extra-vascular manifestations. RESULTS: Fourteen patients with perivascular cuffing were retrieved: three with celiac and superior mesenteric artery (SMA) perivascular cuffing as the only manifestation of surgically proven PDAC, seven with abdominal vasculitis, and four with retroperitoneal fibrosis. PDAC patients exhibited perivascular cuffing of either or both celiac and SMA (3/3). Vasculitis patients showed aortitis with or without iliac or SMA cuffing (3/7) or cuffing of either or both celiac and SMA (4/7). Retroperitoneal fibrosis involved the aorta (4/4), common iliac (4/4), and renal arteries (2/4). Hydronephrosis was present in 3/4 of retroperitoneal fibrosis patients. PDAC and vasculitis demonstrated reduced wall thickness in comparison to retroperitoneal fibrosis (PDAC: 1.0 ± 0.2 cm, vasculitis: 1.2 ± 0.5 cm, retroperitoneal fibrosis: 2.4 ± 0.4 cm; P = 0.002). There was no significant difference in length of vascular involvement (PDAC: 6.3 ± 2.1 cm, vasculitis: 7.1 ± 2.6 cm, retroperitoneal fibrosis: 8.7 ± 0.5 cm). CONCLUSIONS: Celiac and SMA perivascular cuffing can be the sole finding in PDAC and may be indistinguishable from vasculitis. This entity may differ from retroperitoneal fibrosis as it spares the aorta, iliac, and renal arteries and demonstrates thinner walls and no hydronephrosis.

Undetected Pancreatic Adenocarcinoma on Computed Tomography: Frequency According to Scan Protocol.

BACKGROUND: Computed tomography (CT) is the main diagnostic modality for detecting pancreatic adenocarcinoma. OBJECTIVES: To assess the frequency of missed pancreatic adenocarcinoma on CT scans according to different CT protocols. METHODS: The medical records of consecutive pancreatic adenocarcinoma patients were retrospectively collected (12/2011-12/2015). Patients with abdominal CT scans performed up to a year prior to cancer diagnosis were included. Two radiologists registered the presence of radiological signs of missed cancers. The frequency of missed cancers was compared between portal and pancreatic/triphasic CT protocols. RESULTS: Overall, 180 CT scans of pancreatic adenocarcinoma patients performed prior to cancer diagnosis were retrieved; 126/180 (70.0%) were conducted using pancreatic/triphasic protocols and 54/180 (30.0%) used portal protocols. The overall frequency of missed cancers was 6/180 (3.3%) in our study population. The frequency of missed cancers was higher with the portal CT protocols compared to the pancreatic/triphasic protocols: 5/54 (9.3%) vs. 1/126 (0.8%), P = 0.01. CT signs of missed cancers included small hypodense lesions, peri-pancreatic fat stranding, and dilated pancreatic duct with a cut-off sign. CONCLUSIONS: The frequency of missed pancreatic adenocarcinoma is higher on portal CT protocols. Physicians should consider the cancer miss rate on different CT protocols.

FDG PET/CT radiomics as a tool to differentiate between reactive axillary lymphadenopathy following COVID-19 vaccination and metastatic breast cancer axillary lymphadenopathy: a pilot study.

OBJECTIVES: To evaluate if radiomics with machine learning can differentiate between F-18-fluorodeoxyglucose (FDG)-avid breast cancer metastatic lymphadenopathy and FDG-avid COVID-19 mRNA vaccine-related axillary lymphadenopathy. MATERIALS AND METHODS: We retrospectively analyzed FDG-positive, pathology-proven, metastatic axillary lymph nodes in 53 breast cancer patients who had PET/CT for follow-up or staging, and FDG-positive axillary lymph nodes in 46 patients who were vaccinated with the COVID-19 mRNA vaccine. Radiomics features (110 features classified into 7 groups) were extracted from all segmented lymph nodes. Analysis was performed on PET, CT, and combined PET/CT inputs. Lymph nodes were randomly assigned to a training (n = 132) and validation cohort (n = 33) by 5-fold cross-validation. K-nearest neighbors (KNN) and random forest (RF) machine learning models were used. Performance was evaluated using an area under the receiver-operator characteristic curve (AUC-ROC) score. RESULTS: Axillary lymph nodes from breast cancer patients (n = 85) and COVID-19-vaccinated individuals (n = 80) were analyzed. Analysis of first-order features showed statistically significant differences (p < 0.05) in all combined PET/CT features, most PET features, and half of the CT features. The KNN model showed the best performance score for combined PET/CT and PET input with 0.98 (± 0.03) and 0.88 (± 0.07) validation AUC, and 96% (± 4%) and 85% (± 9%) validation accuracy, respectively. The RF model showed the best result for CT input with 0.96 (± 0.04) validation AUC and 90% (± 6%) validation accuracy. CONCLUSION: Radiomics features can differentiate between FDG-avid breast cancer metastatic and FDG-avid COVID-19 vaccine-related axillary lymphadenopathy. Such a model may have a role in differentiating benign nodes from malignant ones. KEY POINTS: • Patients who were vaccinated with the COVID-19 mRNA vaccine have shown FDG-avid reactive axillary lymph nodes in PET-CT scans. • We evaluated if radiomics and machine learning can distinguish between FDG-avid metastatic axillary lymphadenopathy in breast cancer patients and FDG-avid reactive axillary lymph nodes. • Combined PET and CT radiomics data showed good test AUC (0.98) for distinguishing between metastatic axillary lymphadenopathy and post-COVID-19 vaccine-associated axillary lymphadenopathy. Therefore, the use of radiomics may have a role in differentiating between benign from malignant FDG-avid nodes.

Artificial intelligence for identification of focal lesions in intraoperative liver ultrasonography.

PURPOSE: Intraoperative ultrasonography (IOUS) of the liver is a crucial adjunct in every liver resection and may significantly impact intraoperative surgical decisions. However, IOUS is highly operator dependent and has a steep learning curve. We describe the design and assessment of an artificial intelligence (AI) system to identify focal liver lesions in IOUS. METHODS: IOUS images were collected during liver resections performed between November 2020 and November 2021. The images were labeled by radiologists and surgeons as normal liver tissue versus images that contain liver lesions. A convolutional neural network (CNN) was trained and tested to classify images based on the labeling. Algorithm performance was tested in terms of area under the curves (AUCs), accuracy, sensitivity, specificity, F1 score, positive predictive value, and negative predictive value. RESULTS: Overall, the dataset included 5043 IOUS images from 16 patients. Of these, 2576 were labeled as normal liver tissue and 2467 as containing focal liver lesions. Training and testing image sets were taken from different patients. Network performance area under the curve (AUC) was 80.2 ± 2.9%, and the overall classification accuracy was 74.6% ± 3.1%. For maximal sensitivity of 99%, the classification specificity is 36.4 ± 9.4%. CONCLUSIONS: This study provides for the first time a proof of concept for the use of AI in IOUS and show that high accuracy can be achieved. Further studies using high volume data are warranted to increase accuracy and differentiate between lesion types.

Alerting on mortality among patients discharged from the emergency department: a machine learning model.

OBJECTIVES: Physicians continuously make tough decisions when discharging patients. Alerting on poor outcomes may help in this decision. This study evaluates a machine learning model for predicting 30-day mortality in emergency department (ED) discharged patients. METHODS: We retrospectively analysed visits of adult patients discharged from a single ED (1/2014-12/2018). Data included demographics, evaluation and treatment in the ED, and discharge diagnosis. The data comprised of both structured and free-text fields. A gradient boosting model was trained to predict mortality within 30 days of release from the ED. The model was trained on data from the years 2014-2017 and validated on data from the year 2018. In order to reduce potential end-of-life bias, a subgroup analysis was performed for non-oncological patients. RESULTS: Overall, 363 635 ED visits of discharged patients were analysed. The 30-day mortality rate was 0.8%. A majority of the mortality cases (65.3%) had a known oncological disease. The model yielded an area under the curve (AUC) of 0.97 (95% CI 0.96 to 0.97) for predicting 30-day mortality. For a sensitivity of 84% (95% CI 0.81 to 0.86), this model had a false positive rate of 1:20. For patients without a known malignancy, the model yielded an AUC of 0.94 (95% CI 0.92 to 0.95). CONCLUSIONS: Although not frequent, patients may die following ED discharge. Machine learning-based tools may help ED physicians identify patients at risk. An optimised decision for hospitalisation or palliative management may improve patient care and system resource allocation.

Ultra-Low Dose Chest CT with Denoising for Lung Nodule Detection.

BACKGROUND: Medical imaging and the resultant ionizing radiation exposure is a public concern due to the possible risk of cancer induction. OBJECTIVES: To assess the accuracy of ultra-low-dose (ULD) chest computed tomography (CT) with denoising versus normal dose (ND) chest CT using the Lung CT Screening Reporting and Data System (Lung-RADS). METHODS: This prospective single-arm study comprised 52 patients who underwent both ND and ULD scans. Subsequently AI-based denoising methods were applied to produce a denoised ULD scan. Two chest radiologists independently and blindly assessed all scans. Each scan was assigned a Lung-RADS score and grouped as 1 + 2 and 3 + 4. RESULTS: The study included 30 men (58%) and 22 women (42%); mean age 69.9 ± 9 years (range 54-88). ULD scan radiation exposure was comparable on average to 3.6-4.8% of the radiation depending on patient BMI. Denoising increased signal-to-noise ratio by 27.7%. We found substantial inter-observer agreement in all scans for Lung-RADS grouping. Denoised scans performed better than ULD scans when negative likelihood ratio (LR-) was calculated (0.04--0.08 vs. 0.08-0.12). Other than radiation changes, diameter measurement differences and part-solid nodules misclassification as a ground-glass nodule caused most Lung-RADS miscategorization. CONCLUSIONS: When assessing asymptomatic patients for pulmonary nodules, finding a negative screen using ULD CT with denoising makes it highly unlikely for a patient to have a pulmonary nodule that requires aggressive investigation. Future studies of this technique should include larger cohorts and be considered for lung cancer screening as radiation exposure is radically reduced.

Breast MRI during lactation: effects on tumor conspicuity using dynamic contrast-enhanced (DCE) in comparison with diffusion tensor imaging (DTI) parametric maps.

PURPOSE: To investigate the effect of lactation on breast cancer conspicuity on dynamic contrast-enhanced (DCE) MRI in comparison with diffusion tensor imaging (DTI) parametric maps. MATERIALS AND METHODS: Eleven lactating patients with 16 biopsy-confirmed pregnancy-associated breast cancer (PABC) lesions were prospectively evaluated by DCE and DTI on a 1.5-T MRI for pre-treatment evaluation. Additionally, DCE datasets of 16 non-lactating age-matched breast cancer patients were retrospectively reviewed, as control. Contrast-to-noise ratio (CNR) comprising two regions of interests of the normal parenchyma was used to assess the differences in the tumor conspicuity on DCE subtraction images between lactating and non-lactating patients, as well as in comparison against DTI parametric maps of λ1, λ2, λ3, mean diffusivity (MD), fractional anisotropy (FA), and maximal anisotropy index, λ1-λ3. RESULTS: CNR values of breast cancer on DCE MRI among lactating patients were reduced by 62% and 58% (p < 0.001) in comparison with those in non-lactating patients, when taking into account the normal contralateral parenchyma and an area of marked background parenchymal enhancement (BPE), respectively. Among the lactating patients, DTI parameters of λ1, λ2, λ3, MD, and λ1-λ3 were significantly decreased, and FA was significantly increased in PABC, relative to the normal lactating parenchyma ROIs. When compared against DCE in the lactating cohort, the CNR on λ1, λ2, λ3, and MD was significantly superior, providing up to 138% more tumor conspicuity, on average. CONCLUSION: Breast cancer conspicuity on DCE MRI is markedly reduced during lactation owing to the marked BPE. However, the additional application of DTI can improve the visualization and quantitative characterization of PABC, therefore possibly suggesting an additive value in the diagnostic workup of PABC. KEY POINTS: • Breast cancer conspicuity on DCE MRI has decreased by approximately 60% among lactating patients compared with non-lactating controls. • DTI-derived diffusion coefficients and the anisotropy indices of PABC lesions were significantly different than those of the normal lactating fibroglandular tissue. • Among lactating patients, breast cancer conspicuity on DTI-derived parametric maps provided up to 138% increase in contrast-to-noise ratio compared with DCE imaging.

Comparison of deep learning models for natural language processing-based classification of non-English head CT reports.

PURPOSE: Natural language processing (NLP) can be used for automatic flagging of radiology reports. We assessed deep learning models for classifying non-English head CT reports. METHODS: We retrospectively collected head CT reports (2011-2018). Reports were signed in Hebrew. Emergency department (ED) reports of adult patients from January to February for each year (2013-2018) were manually labeled. All other reports were used to pre-train an embedding layer. We explored two use cases: (1) general labeling use case, in which reports were labeled as normal vs. pathological; (2) specific labeling use case, in which reports were labeled as with and without intra-cranial hemorrhage. We tested long short-term memory (LSTM) and LSTM-attention (LSTM-ATN) networks for classifying reports. We also evaluated the improvement of adding Word2Vec word embedding. Deep learning models were compared with a bag-of-words (BOW) model. RESULTS: We retrieved 176,988 head CT reports for pre-training. We manually labeled 7784 reports as normal (46.3%) or pathological (53.7%), and 7.1% with intra-cranial hemorrhage. For the general labeling, LSTM-ATN-Word2Vec showed the best results (AUC = 0.967 ± 0.006, accuracy 90.8% ± 0.01). For the specific labeling, all methods showed similar accuracies between 95.0 and 95.9%. Both LSTM-ATN-Word2Vec and BOW had the highest AUC (0.970). CONCLUSION: For a general use case, word embedding using a large cohort of non-English head CT reports and ATN improves NLP performance. For a more specific task, BOW and deep learning showed similar results. Models should be explored and tailored to the NLP task.

Promoting head CT exams in the emergency department triage using a machine learning model.

PURPOSE: In this study, we aimed to develop a novel prediction model to identify patients in need of a non-contrast head CT exam during emergency department (ED) triage. METHODS: We collected data of all adult ED visits in our institution for five consecutive years (1/2013-12/2017). Retrieved variables included the following: demographics, mode of arrival to the ED, comorbidities, home medications, structured and unstructured chief complaints, vital signs, pain scale score, emergency severity index, ED wing assignment, documentation of previous ED visits, hospitalizations and CTs, and current visit non-contrast head CT usage. A machine learning gradient boosting model was trained on data from the years 2013-2016 and tested on data from 2017. Area under the curve (AUC) was used as metrics. Single-variable AUCs were also determined. Youden's index evaluated optimal sensitivity and specificity of the models. RESULTS: The final cohort included 595,561 ED visits. Non-contrast head CT usage rate was 11.8%. Each visit was coded into an input vector of 171 variables. Single-variable analysis showed that chief complaint had the best single predictive analysis (AUC = 0.87). The best model showed an AUC of 0.93 (95% CI 0.931-0.936) for predicting non-contrast head CT usage at triage level. The model had a sensitivity of 88.1% and specificity of 85.7% for non-contrast head CT utilization. CONCLUSION: The developed model can identify patients that need to undergo head CT exam already in the ED triage level and by that allow faster diagnosis and treatment.

Noncontrast Breast MRI During Pregnancy Using Diffusion Tensor Imaging: A Feasibility Study.

BACKGROUND: Pregnancy-associated breast cancer (PABC) is often a delayed diagnosis and contrast-enhanced MRI is contraindicated because gadolinium agents are known to cross the placenta. PURPOSE: To investigate the feasibility and clinical utility of noncontrast breast MRI using diffusion tensor imaging (DTI) in the diagnostic workup of PABC. STUDY TYPE: Prospective. POPULATION: Between November 2016 and January 2018, 25 pregnant participants (median gestational age: 17 weeks) were recruited from eight referral breast-care centers nationwide. Imaging indications included: newly-diagnosed PABC (n = 10, with 11 lesions), palpable mass/mastitis (n = 4), high-risk screening (n = 10), and monitoring neoadjuvant-chemotherapy response (n = 1). FIELD STRENGTH/SEQUENCE: 1.5T, T2 -weighted, and DTI sequences, prone position, with a scan duration of ∼12 minutes. ASSESSMENT: DTI parametric maps were generated and analyzed at pixel resolution, with reference to ultrasound (US) and pathology. STATISTICAL TESTS: Two-tailed Student's t-test was applied for evaluating differences between DTI parameters of PABC vs. healthy fibroglandular tissue. Pearson's correlation test was applied to measure the agreements between λ1-based longest tumor diameter, US, and pathology. RESULTS: All scans were technically completed and reached diagnostic quality, except one with notable motion artifacts due to positional discomfort, which was excluded. Nine out of 11 known PABC lesions and one newly-diagnosed lesion were visible on λ1, λ2, λ3, mean diffusivity (MD), and λ1-λ3 maps, with substantial parametric contrast compared with the apparently normal contralateral fibroglandular tissue (P < 0.001 for all). Two lesions of 0.7 cm were not depicted by the diffusivity maps. Tumor diameter measured on a thresholded λ1 map correlated well with US (r = 0.97) and pathology (r = 0.95). Malignancy was excluded by DTI parametric maps in scans of symptomatic and high-risk patients, in agreement with US follow-up, except for one false-positive case. DATA CONCLUSION: Noncontrast breast MRI is feasible and well-tolerated during pregnancy. Further studies with a larger and homogeneous cohort are required to validate DTI's additive diagnostic value, albeit this study suggests a potential adjunct role for this noninvasive approach in breast evaluation during pregnancy. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:508-517.

Application of MR morphologic, diffusion tensor, and perfusion imaging in the classification of brain tumors using machine learning scheme.

PURPOSE: While MRI is the modality of choice for the assessment of patients with brain tumors, differentiation between various tumors based on their imaging characteristics might be challenging due to overlapping imaging features. The purpose of this study was to apply a machine learning scheme using basic and advanced MR sequences for distinguishing different types of brain tumors. METHODS: The study cohort included 141 patients (41 glioblastoma, 38 metastasis, 50 meningioma, and 12 primary central nervous system lymphoma). A computer-assisted classification scheme, combining morphologic MRI, perfusion MRI, and DTI metrics, was developed and used for tumor classification. The proposed multistep scheme consists of pre-processing, ROI definition, features extraction, feature selection, and classification. Feature subset selection was performed using support vector machines (SVMs). Classification performance was assessed by leave-one-out cross-validation. Given an ROI, the entire classification process was done automatically via computer and without any human intervention. RESULTS: A binary hierarchical classification tree was chosen. In the first step, selected features were chosen for distinguishing glioblastoma from the remaining three classes, followed by separation of meningioma from metastasis and PCNSL, and then to discriminate PCNSL from metastasis. The binary SVM classification accuracy, sensitivity and specificity for glioblastoma, metastasis, meningiomas, and primary central nervous system lymphoma were 95.7, 81.6, and 91.2%; 92.7, 95.1, and 93.6%; 97, 90.8, and 58.3%; and 91.5, 90, and 96.9%, respectively. CONCLUSION: A machine learning scheme using data from anatomical and advanced MRI sequences resulted in high-performance automatic tumor classification algorithm. Such a scheme can be integrated into clinical decision support systems to optimize tumor classification.

CT measurement of breast glandular tissue and its association with testicular cancer.

OBJECTIVES: To evaluate the associations between breast glandular tissues diameters as determined by CT and b-hCG levels, histological types, tumour spread and prognosis in patients with testicular germ cell tumour. METHODS: Ninety-four patients with pre-treatment CT scan and markers (b-hCG, AFP, LDH) were retrospectively collected. A radiologist measured diameters in all CT examinations and correlation between diameters and log (b-hCG) was assessed (Pearson's coefficient). The ability of measured diameters to predict lymphatic and distant haematogenous metastatic spread was evaluated (ROC curves). The associations between measured diameter cut-off values of 20 and 25 mm and International Germ Cell Cancer Collaborative Group (IGCCCG) classification, lymphatic and distant haematogenous metastatic spread and histological subtypes were evaluated (chi squared test). RESULTS: Breast glandular diameters correlated to log(b-hCG) (r = 0.579) and predicted distant haematogenous metastatic spread (AUC = 0.78). Worse prognosis (intermediate or poor IGCCCG) was shown for 20 mm (27.3 vs. 4.2 %, p = 0.005) and 25 mm (33.3 vs. 6.1 %, p = 0.014). A diameter of 25 mm was associated with non-seminoma (91.7 vs. 48.8 %, p = 0.005). CONCLUSION: Breast glandular tissue diameters correlated with log(b-hCG) and predicted distant haematogenous metastases. Twenty and 25 mm were associated with worse prognosis and 25 mm was able to distinguish between seminoma and non-seminoma. KEY POINTS: • CT breast glandular tissue diameter correlates with log(b-HCG) • Gynaecomastia in CT is associated with worse prognosis • Gynaecomastia in CT is associated with non-seminoma histological subtype.

Intraluminal Uterine Hypodensity in CT Scans of Postmenopausal Women: Recommendations for Interpretation.

OBJECTIVE: This study aims to investigate the association between intraluminal uterine hypodensity and uterine malignancy and establish thresholds that would minimize routine gynecological evaluation. METHODS: Two groups were recruited retrospectively: cancer group, which comprised 32 sequential endometrial cancer patients, and postmenopausal group, which comprised 63 women, with no known gynecologic malignancy.Two radiologists independently measured hypodensity, transversely in the axial plane and anterioposteriorly in the sagittal plane.The association between cancer and hypodensity was evaluated. Receiver operating characteristic curves were evaluated diameters predictive of cancer. RESULTS: Hypodensity was associated with cancer (cancer group, 93.8% vs. postmenopausal group, 38.1%; P < 0.0001). Hypodensity diameters correlated highly with prediction of cancer (transverse area under the curve, 0.899; anteroposterior area under the curve, 0.892). Diameters of 19.5 mm transverse and 6.0 mm anteroposterior yielded a sensitivity of 87% and 83% and specificity of 91% and 83%, respectively. CONCLUSIONS: Intrauterine hypodensity is a common finding in computed tomography scans of postmenopausal women. A transverse diameter of 19.5 mm and an anteroposterior diameter of 6.0 mm are suggested as thresholds for further gynecological sonographic evaluation.

Role of Emergency Magnetic Resonance Imaging for the Workup of Suspected Appendicitis in Pregnant Women.

BACKGROUND: Pregnant women with acute abdominal pain pose a diagnostic challenge. Delay in diagnosis may result in significant risk to the fetus. The preferred diagnostic modality is magnetic resonance imaging (MRI), since ultrasonography is often inconclusive, and computed tomography (CT) would expose the fetus to ionizing radiation. OBJECTIVES: To describe the process in setting up an around-the-clock MRI service for diagnosing appendicitis in pregnant women and to evaluate the contribution of abdominal MR in the diagnosis of acute appendicitis. METHODS: We conducted a retrospective study of consecutive pregnant women presenting with acute abdominal pain over a 6 year period who underwent MRI studies. A workflow that involved a multidisciplinary team was developed. A modified MRI protocol adapted to pregnancy was formulated. Data regarding patients' characteristics, imaging reports and outcome were collected retrospectively. RESULTS: 49 pregnant women with suspected appendicitis were enrolled. Physical examination was followed by ultrasound: when positive, the patients were referred for MR scan or surgery treatment; when the ultrasound was inconclusive, MR scan was performed. In 88% of women appendicitis was ruled out and surgery was prevented. MRI diagnosed all cases with acute appendicitis and one case was inconclusive. The overall statistical performance of the study shows a negative predictive value of 100% (95%CI 91.9-100%) and positive predictive value of 83.3% (95%CI 35.9-99.6%). CONCLUSIONS: Creation of an around-the-clock imaging service using abdominal MRI with the establishment of a workflow chart using a dedicated MR protocol is feasible. It provides a safe way to rule out appendicitis and to avoid futile surgery in pregnant women.