The Role of Whole-Body Magnetic Resonance Imaging in Assessing Extrapulmonary Metastases in Osteosarcoma Staging and Restaging: A Pilot Study.

OBJECTIVE: The aim of the study is to investigate the role of whole-body magnetic resonance imaging (MRI) in assessing extrapulmonary metastases in primary osteosarcoma staging. METHODS: We retrospectively reviewed medical data to identify primary osteosarcoma patients with available preoperative whole-body MRI obtained in the staging or restaging. Histopathology was the reference test for assessing the diagnostic performance, if available. Otherwise, oncology board decisions were used as the reference. In addition, the benefits of whole-body MRI to F-fluorodeoxyglucose positron emission tomography-computed tomography (18F-FDG PET-CT) and bone scintigraphy were investigated. RESULTS: In all, 36 patients with osteosarcoma (24 staging, 12 restaging) with a mean age of 16.36 ± 5.63 years (range, 9-29 years) were included in the study. The median follow-up duration was 26.61 months (interquartile range, 33.3 months). Of 36 patients, 8 had skeletal, 1 had a lymph node, and 1 had a subcutaneous metastasis. Whole-body MRI correctly identified all patients with metastatic disease but incorrectly classified a bone infarct in one patient as a skeletal metastasis, equating a scan-level sensitivity, specificity, accuracy, negative predictive value, and positive predictive value of 100%, 96.3%, 97.3%, 100%, and 90.91%. Whole-body MRI contributed to bone scintigraphy by identifying a skeletal metastasis in one patient and positron emission tomography-computed tomography by ruling out a skeletal metastasis in another. CONCLUSIONS: Whole-body MRI could accurately identify extrapulmonary metastases in primary osteosarcoma patients for staging or restaging. In addition, it might contribute to the standard whole-body imaging methods.

The analysis of 18 F-FDG PET/MRI, electroencephalography, and semiology in patients with gray matter heterotopia: A pilot study.

OBJECTIVES: To describe 18 F-fluorodeoxyglucose positron emission tomography/magnetic resonance imaging (18 F-FDG PET/MRI) along with semiology and electroencephalography (EEG) in patients with gray matter heterotopia (GMH); to evaluate the concordance between 18 F-FDG PET/MRI and clinical epileptogenic zone (EZ). MATERIALS & METHODS: GMH (subcortical heterotopia [SCH] and periventricular nodular heterotopia [PNH]) patients with epilepsy who underwent 18 F-FDG PET/MRI were retrospectively enrolled. Two radiologists evaluated brain MRI, while two nuclear medicine specialists assessed the 18 F-FDG PET. The SUVmax values of visually hypometabolic cortical areas were compared to the contralateral cortex using a SUVmax threshold value of 10%; the SUVmax values of GMH lesions were compared with that of the right precentral gyrus. The cortex or GMH with hypometabolism on 18 F-FDG PET/MRI was considered representative of the EZ. The clinical EZ was identified using EEG and semiology. RESULTS: Thirty patients (19 PNH; 11 SCH) with a mean age of 28.46 ± 9.52 years were enrolled. The heterotopic nodules were ametabolic in 3 patients (10%), hypometabolic in 16 (33.33%), isometabolic in 13 (26.66%), and hypermetabolic in 4 (10%). 18 F-FDG PET/MRI demonstrated hypometabolism in the cortex and GMH in 22/30 (73.33%) and 16/30 (53.33%). We could identify a clinical EZ in 18 patients, and 15 out of 18 (83.33%) had concordant 18 F-FDG PET/MRI findings. CONCLUSION: Heterotopic nodules in GMH patients show different metabolic patterns on 18 F-FDG PET/MRI, with nearly three-quarters of the patients having cortical hypometabolism. 18 F-FDG PET/ MRI findings are mostly concordant with the clinical EZ.

Third-Generation Dual-Source Computed Tomography for Coronary Angiography With Individually Tailored Scan Protocols Can Achieve a Low Radiation Dose With Good Image Quality in Unselected Patients.

OBJECTIVE: To explore the image quality and radiation exposure associated with coronary angiography obtained with a third-generation dual-source computed tomography, using body mass index (BMI)- and heart rate (HR)-adapted protocols in real-world patients. METHODS: Three scan protocols were implemented with regard to HR: prospective turbo high-pitch spiral, sequential, and retrospective spiral modes. We adapted the reference kilovoltage value according to BMI. Image quality was evaluated using a 4-point scale, and effective dose estimates were calculated using the dose-length product. RESULTS: Among the 896 patients, 417 (46.54%), 433 (48.32%), and 45 (5.02%) were imaged using prospective turbo high-pitch spiral, sequential, and retrospective spiral modes, respectively. The median BMI was 27.3 (25-30.4) kg/m2, and the effective dose was 0.65 mSv (interquartile range, 0.33-1.56 mSv). Only 32 of 896 examinations (3.5%) had poor image quality. CONCLUSIONS: Computed tomography angiography with BMI- and HR-tailored protocols offers good image quality with low radiation dose in unselected patients.

The influence of cardiac motion on radiomics features: radiomics features of non-enhanced CMR cine images greatly vary through the cardiac cycle.

OBJECTIVES: The cardiac cycle might impair the reproducibility of radiomics features of cardiac magnetic resonance (CMR) cine images, yet this issue has not been addressed in the previous research. We aim to evaluate whether radiomics features of CMR cine images vary during the cardiac cycle and investigate the reproducibility of radiomics features of CMR cine images. METHODS: This retrospective study enrolled 59 healthy adults who underwent CMR examination. Two observers segmented the myocardium on a 4D stack of three consecutive mid-ventricular short-axis cine images covering the cardiac cycle. A total of 352 radiomics features were extracted. The coefficient of variation and intraclass correlation coefficient were used to assess the feature variability through the cycle and inter-observer reproducibility, respectively. RESULTS: Approximately 55% of radiomics features showed large variability through the cardiac cycle. The original features showed more variability than the Laplacian of Gaussian-filtered features (73.8% vs. 48%). The features of 4D stack cine images had a higher proportion of reproducible features (92.0%, 87.7%, and 76.1%) compared with the end-diastolic (77.8%, 62.2%, and 41.7%) and the end-systolic images (81.5%, 74.1%, and 58.8%) for intraclass correlation cut-off values of 30.80, > 0.85, and > 0.90, respectively. CONCLUSIONS: Radiomics features of CMR cine images greatly vary during the cardiac cycle. The radiomics features of 4D stack of cine images are more robust compared with end-diastolic and end-systolic cine images in terms of reproducibility. The impact of the cardiac cycle on the reproducibility of the features should be considered when employing CMR cine images radiomics. KEY POINTS: • There is limited evidence on the impact of cardiac motion on radiomics features of CMR cine images and the reproducibility of the radiomics features of CMR cine images. • Radiomics features of non-enhanced CMR cine images greatly vary during the cardiac cycle, and the number of "reproducible" features shows significant variations according to the cardiac phases. • The impact of cardiac cycle on the reproducibility of the radiomics features should be considered when employing CMR cine images radiomics.

The diagnostic contribution of intracranial vessel wall imaging in the differentiation of primary angiitis of the central nervous system from other intracranial vasculopathies.

PURPOSE: The aim of this study is to demonstrate the diagnostic effect of VWI in differentiating PACNS from other vasculopathies and its role in post-treatment follow-up in PACNS patients in this study. METHODS: In this prospective study, we included patients with clinical suspicion of PACNS who presented with new-onset ischemic events and had significant intracranial large vessel stenosis on DSA or MRA. VWI was performed on all patients. The imaging findings and final diagnoses were recorded. Control VWI was performed on patients with PACNS diagnosis after at least 3 months of treatment, and the change in findings was also evaluated. RESULTS: Twenty-three patients were included in the study had a median age of 40 (range 12-58). The most common clinical manifestations were focal neurologic deficits. According to the initial clinical evaluation, 10 patients (43.5%) were classified as PACNS and 13 patients (56.5%) as indeterminate for PACNS. After incorporating the VWI findings, the diagnosis of PACNS was confirmed in all clinically diagnosed PACNS patients. Concentric wall thickening and contrast enhancement were statistically significant in the PACNS group (p <0.001). According to concentric thickening and VWE features, sensitivity and specificity in distinguishing PACNS and other vasculopathies were 95.2%, 75% and 95.2%, 68.8%, respectively. Vessel wall enhancement regressed in 7 of 9 patients during a median follow-up period of 8 months (range 5.5-11.5) in PACNS patients who followed up. CONCLUSION: VWI seems a new and useful imaging method in the differential diagnosis of PACNS and might be a useful adjunct for post-treatment follow-up.

Comparison of MRI, PSMA PET/CT, and Fusion PSMA PET/MRI for Detection of Clinically Significant Prostate Cancer.

PURPOSE: The aim of our study is to compare the efficacy of positron emission tomography (PET) and magnetic resonance imaging (MRI) for detecting intraprostatic lesions in patients with clinically significant prostate cancer who underwent radical prostatectomy; additionally, investigate the benefits of rostate-specific membrane antigen (PSMA) PET-MR software fusion images to the diagnosis. METHODS: Thirty patients, who underwent radical prostatectomy between June 2015 and April 2018, were included in the study. Subjects with gallium PSMA PET-CT and multiparametric prostate MRI performed according to Prostate Imaging Reporting and Data System v2 criteria in our clinic were included in the study. 68Ga-PSMA PET-CT images were fused with MR sequences for analysis. RESULTS: The mean age of cases was 63.2 years (ranged from 45 to 79 years). Index lesions of 29 cases were detected by MRI and 22 of them by PET CT. Both modalities were found to be less sensitive for detection of bilaterality and multifocality (42.85% and 20% for MRI, 28.57% and 20% for PET CT, respectively). There was no statistically significant difference between modalities. It was observed that if a clinically significant tumor focus was not detected by MRI, it was small (6 mm or less) in diameter or had a low Gleason score. CONCLUSIONS: Software fusion PSMA PET-MRI increased the sensitivity of the index lesion identification compared with PSMA PET-CT and also increased the sensitivity of real lesion size identification compared with multiparametric prostate MRI.

Benign leiomyoma with multiple metastases to vertebrae and calvarium: An index case with comprehensive review of endocrine targets.

"Benign" metastatic leiomyomas (BML) are indolently growing metastatic tumors which mostly associate with uterine leiomyomas in women in reproductive ages. The reason to define these lesions as "benign" despite metastasis is their pathological features with low mitotic counts, lack of or minimal nuclear atypia, pseudocyst formation, and coagulative necrosis unlike leiomyosarcomas. Despite lack of pathological malignant features, they may cause significant morbidity and even mortality. Here, we describe a BML case with metastases to vertebrae and skull bones. Vertebral and skull metastases of BMLs were very rarely reported. In treatment of these tumors, hysterectomy and GnRH modifier treatments are widely employed. GnRH agonists act by desensitization and downregulation of the GnRH receptors, while GnRH antagonists act via the canonical competitive blockage. These treatments reduce FSH and LH levels, thereby reducing the systemic levels of sex steroids which stimulate leiomyoma growth. However, leiomyomas inherently harbor aromatase activity and synthesize their own estrogen; hence, treatment with systemic estrogen antagonists may provide better tumor control. Another important factor in BML pathogenesis is progesterone, and both progesterone receptor antagonists and high-dose progesterone receptor agonists may reduce BML growth. Following surgical treatment of the calvarial mass and radiotherapy of the vertebral metastatic foci, our BML case was successfully managed with hysterectomy and anastrozole treatment. Higher awareness of BML cases and their molecular endocrinological features in the neurosurgical community may pave to develop better strategies for treatment of these tumors causing high morbidity.

Diagnostic value of left atrial strain in pediatric hypertrophic cardiomyopathy with normal maximum left atrial volume index: preliminary cardiac magnetic resonance study.

BACKGROUND: The maximum left atrial volume index is the most widely used metric for assessing the left atrium in hypertrophic cardiomyopathy; however, it may be normal in the early phases of the disease. OBJECTIVE: To assess whether pediatric hypertrophic cardiomyopathy patients with normal maximum left atrial volume index have impaired atrial functions on cardiac magnetic resonance imaging (MRI). MATERIALS AND METHOD: A total of 26 pediatric hypertrophic cardiomyopathy patients and 24 age-matched children, as controls, were enrolled in the study. The left atrial reservoir, conduit and booster strain were calculated from two orthogonal planes and the left atrial volumes were calculated using the biplanar method. The extent of left ventricular late gadolinium enhancement (LGE-%) was calculated using the thresholding method. The left ventricular early diastolic longitudinal strain rate was calculated to assess diastolic dysfunction. RESULTS: The maximum left atrial volume index of the children with hypertrophic cardiomyopathy and the controls were not significantly different (P>0.05). Most of the left atrial functional indices were worse in children with hypertrophic cardiomyopathy (P<0.05), yet no difference was observed between the left atrial booster strains of the two groups (P>0.05). The left atrial conduit strain showed moderate to good negative correlations with left ventricular LGE-% and diastolic dysfunction. CONCLUSION: Left atrial conduit and reservoir strains are impaired in pediatric hypertrophic cardiomyopathy patients with normal maximum left atrial volumetric indices. Most left atrial strain parameters are closely linked to left ventricular LGE-% and diastolic dysfunction. Left atrial strain analysis may reveal subtle functional changes in the atrium before the increase in the maximum volume index.

Does Employing a Flowchart Improve the Diagnostic Performance of Cardiac Magnetic Resonance Imaging in Left Ventricular Noncompaction?

BACKGROUND: To test the hypothesis that making a diagnosis of left ventricular noncompaction (LVNC) on cardiac magnetic resonance imaging (CMRI) using a noncompacted-to-compacted (NC/C) myocardium ratio > 2.3 would yield significant errors, and also to test a diagnostic flowchart in patients who undergo CMRI and have clinical and echocardiographic findings suggesting LVNC could improve the diagnosis of LVNC. METHODS: A total of 84 patients with LVNC and 162 controls consisting of patients with other diseases and healthy participants who had CMRI and echocardiograms were selected. The diagnostic flowchart of the study involved the use of CMRI with all available sequences for patients with a high pre-test probability of LVNC. Two blinded independent cardiologists evaluated echocardiograms, and patients with suggestive echocardiographic and clinical findings for LVNC were enrolled in the high pre-test probability of LVNC group. Two independent blinded radiologists established the diagnosis of LVNC based on NC/C ratio > 2.3 on CMRI, and they were allowed to re-assess the patients following the diagnostic flowchart. RESULTS: An NC/C ratio > 2.3 identified 83 of 84 LVNC patients, yet incorrectly classified 48 of the 162 controls as having LVNC. Radiologists changed their decision in 23 of 48 patients with incorrect diagnoses, resulted in improved specificity (70.4% to 84.6%). The use of the CMRI diagnostic flowchart in the high pre-test probability group yielded a high specificity (97.2%) and accuracy (95.9%). CONCLUSIONS: LVNC diagnosed by CMRI based on the NC/C criterion can lead to overdiagnosis, whereas only using CMRI in patients with a high pre-test probability of LVNC with all available sequences may improve the diagnostic performance.

Brain MRI Findings in Patients in the Intensive Care Unit with COVID-19 Infection.

Online supplemental material is available for this article.

Multicenter Multireader Evaluation of an Artificial Intelligence-Based Attention Mapping System for the Detection of Prostate Cancer With Multiparametric MRI.

OBJECTIVE. The purpose of this study was to evaluate in a multicenter dataset the performance of an artificial intelligence (AI) detection system with attention mapping compared with multiparametric MRI (mpMRI) interpretation in the detection of prostate cancer. MATERIALS AND METHODS. MRI examinations from five institutions were included in this study and were evaluated by nine readers. In the first round, readers evaluated mpMRI studies using the Prostate Imaging Reporting and Data System version 2. After 4 weeks, images were again presented to readers along with the AI-based detection system output. Readers accepted or rejected lesions within four AI-generated attention map boxes. Additional lesions outside of boxes were excluded from detection and categorization. The performances of readers using the mpMRI-only and AI-assisted approaches were compared. RESULTS. The study population included 152 case patients and 84 control patients with 274 pathologically proven cancer lesions. The lesion-based AUC was 74.9% for MRI and 77.5% for AI with no significant difference (p = 0.095). The sensitivity for overall detection of cancer lesions was higher for AI than for mpMRI but did not reach statistical significance (57.4% vs 53.6%, p = 0.073). However, for transition zone lesions, sensitivity was higher for AI than for MRI (61.8% vs 50.8%, p = 0.001). Reading time was longer for AI than for MRI (4.66 vs 4.03 minutes, p < 0.001). There was moderate interreader agreement for AI and MRI with no significant difference (58.7% vs 58.5%, p = 0.966). CONCLUSION. Overall sensitivity was only minimally improved by use of the AI system. Significant improvement was achieved, however, in the detection of transition zone lesions with use of the AI system at the cost of a mean of 40 seconds of additional reading time.

Whole-Body MRI Screening in Asymptomatic Subjects; Preliminary Experience and Long-Term Follow-Up Findings.

BACKGROUND: The aim of this study is to describe the technique and to evaluate the results of whole-body magnetic resonance imaging in an asymptomatic population. MATERIAL/METHODS: Between March 2009 and December 2011, 118 consecutive subjects undergoing thorough medical check-up were prospectively included in the study. MRI was performed with a 205-cm moving table, parallel imaging and automatic image composing software. RESULTS: In 83 subjects (70%), 103 benign lesions were detected. Two malignant (adrenal and renal carcinoma) lesions and one precancerous (pancreatic mucinous carcinoma) lesion were detected. The most common lesions were renal cysts, liver hemangiomas, liver cysts, thyroid nodules, and uterine leiomyomas. CONCLUSIONS: WB-MRI is able to cover area from head to toes in one diagnostic work-up, and besides the anatomic regions evaluated by conventional radiological modalities, i.e. brain parenchyma, bones and extremities, can be evaluated in one examination.

Simple electrodiagnostic method for Morton neuroma.

INTRODUCTION: We describe a simple and quickly applied electrodiagnostic method for confirming the diagnosis of interdigital neuropathy caused by Morton neuroma (MN). METHODS: Interdigital nerves II-III and III-IV were stimulated with surface electrodes simultaneously touching the lateral side of 1 toe and the medial side of the other. Recording was also made with surface electrodes. The results of 20 normal controls and 14 patients with MN were evaluated. RESULTS: The amplitude and peak latency values elicited in the patients as well as the interside differences revealed an acceptable abnormality rate between 57.1% and 71.4%. CONCLUSIONS: Although the most popular and effective method of MN diagnosis is clinical evaluation supported by imaging, electrophysiological studies can, in selected patients, provide valuable information.

Diagnostic value of strain echocardiography, galectin-3, and tenascin-C levels for the identification of patients with pulmonary and cardiac sarcoidosis.

BACKGROUND: Cardiac involvement in sarcoidosis has been associated with poor prognosis. We evaluated myocardial contractility quantitatively in a cohort of pulmonary sarcoidosis (PS) patients with and without cardiac involvement. We also studied markers of fibrosis (tenascin-C [Tn-C] and galectin-3 [Gl-3]) as diagnostic tools for PS and cardiac sarcoidosis (CS). METHODS: Forty ambulatory patients with PS of grades 1-2 and 26 healthy subjects were prospectively enrolled. All patients with PS underwent cardiac magnetic resonance (CMR) to explore the presence of CS. The study population was divided into three groups: controls (n = 26), non-CS patients (n = 34), and CS patients (n = 6). Speckle-tracking strain echocardiography (STE) was performed on all patients, and Gl-3 and Tn-C values were measured in all patients and controls. RESULTS: PS patients had higher levels of Gl-3 and Tn-C than did controls, and the STE parameters of PS patients, including global longitudinal strain (GLS) and global circumferential strain (GCS), were lower than those of controls (p < 0.001 for all comparisons). GLS values were lower in CS patients than in the other groups (p = 0.05). CONCLUSIONS: PS patients demonstrate reduced cardiac contractility, independent of CMR-proven structural cardiac lesions, while patients with structural lesions have a more pronounced drop in strain parameters. Tn-C and Gl-3 are promising markers for the diagnosis of PS, but they are not specific for cardiac involvement.

A new method for computer-assisted detection, definition and differentiation of the urinary calculi.

PURPOSE: Urinary stones are common and can be diagnosed with computed tomography (CT) easily. In this study, we aimed to specify the opacity characteristics of various types of calcified foci that develop through the urinary system by using an image analysis program. With this method, we try to differentiate the calculi from the non-calculous opacities and also we aimed to present how to identify the characteristic features of renal and ureteral calcules. MATERIALS AND METHODS: We obtained the CT studies of the subjects (n = 48, mean age = 41 years) by using a dual source CT imaging system. We grouped the calculi detected in the dual-energy CT sections as renal (n = 40) or ureteric (n = 45) based on their locations. Other radio-opaque structures that were identified outside but within close proximity of the urinary tract were recorded as calculi "mimickers". We used ImageJ program for morphological analysis. All the acquired data were analyzed statistically. RESULTS: According to thorough morphological parameters, there were statistically significant differences in the angle and Feret angle values between calculi and mimickers (p < 0.001). Multivariate logistical regression analysis showed that Minor Axis and Feret angle parameters can be used to distinguish between ureteric (p = 0.003) and kidney (p = 0.001) stones. CONCLUSIONS: Computer-based morphologic parameters can be used simply to differentiate between calcular and noncalcular densities on CT and also between renal and ureteric stones.

Reliability of high-pitch ultra-low-dose paranasal sinus computed tomography for evaluating paranasal sinus anatomy and sinus disease.

OBJECTIVES: The aim of this study was to evaluate the reliability of high-pitch ultra-low-dose computed tomography (CT) for detecting important paranasal sinus anatomic landmarks and pathologies. MATERIALS AND METHODS: Sixty patients (22 females, 38 males) aged 15 to 67 years (mean age, 33.68 y; SD, 9.83 y) underwent high-pitch ultra-low-dose CT of the paranasal sinuses between February and June 2012. To determine the lowest possible dose for evaluation of the paranasal sinuses, the patients were divided into three groups randomly and prospectively. A different low-dose CT protocol was applied to each group. The image quality was assessed subjectively by a radiologist and an otorhinolaryngology head and neck surgeon independently using a 4-point grading scale (0 = structures could not be identified, 1 = indistinctly defined structures, 1.5 = relatively well-defined structures, 2 = very well-defined structures). Anatomic landmarks and mucosal structures were evaluated. Mean scores were evaluated to assess statistical significance. RESULTS: According to the anatomic landmark scoring, excluding the ethmoid foramen for ethmoid artery identification, all of the structures in all 3 groups were very well-defined structures. The ethmoid foramen for ethmoid artery identification was scored as either could not be identified or an indistinctly defined structure in all groups. On evaluating the mucosa of the paranasal sinuses, normal and pathologic mucosal structures were scored as very well defined in all of the patients. The interobserver agreement was excellent. CONCLUSION: High-pitch ultra-low-dose CT is a safe, reliable paranasal sinus screening tool.

Quantitative evaluation of Saliency-Based Explainable artificial intelligence (XAI) methods in Deep Learning-Based mammogram analysis.

BACKGROUND: Explainable Artificial Intelligence (XAI) is prominent in the diagnostics of opaque deep learning (DL) models, especially in medical imaging. Saliency methods are commonly used, yet there's a lack of quantitative evidence regarding their performance. OBJECTIVES: To quantitatively evaluate the performance of widely utilized saliency XAI methods in the task of breast cancer detection on mammograms. METHODS: Three radiologists drew ground-truth boxes on a balanced mammogram dataset of women (n = 1496 cancer-positive and negative scans) from three centers. A modified, pre-trained DL model was employed for breast cancer detection, using MLO and CC images. Saliency XAI methods, including Gradient-weighted Class Activation Mapping (Grad-CAM), Grad-CAM++, and Eigen-CAM, were evaluated. We utilized the Pointing Game to assess these methods, determining if the maximum value of a saliency map aligned with the bounding boxes, representing the ratio of correctly identified lesions among all cancer patients, with a value ranging from 0 to 1. RESULTS: The development sample included 2,244 women (75%), with the remaining 748 women (25%) in the testing set for unbiased XAI evaluation. The model's recall, precision, accuracy, and F1-Score in identifying cancer in the testing set were 69%, 88%, 80%, and 0.77, respectively. The Pointing Game Scores for Grad-CAM, Grad-CAM++, and Eigen-CAM were 0.41, 0.30, and 0.35 in women with cancer and marginally increased to 0.41, 0.31, and 0.36 when considering only true-positive samples. CONCLUSIONS: While saliency-based methods provide some degree of explainability, they frequently fall short in delineating how DL models arrive at decisions in a considerable number of instances.

Choosing the right artificial intelligence solutions for your radiology department: key factors to consider.

The rapid evolution of artificial intelligence (AI), particularly in deep learning, has significantly impacted radiology, introducing an array of AI solutions for interpretative tasks. This paper provides radiology departments with a practical guide for selecting and integrating AI solutions, focusing on interpretative tasks that require the active involvement of radiologists. Our approach is not to list available applications or review scientific evidence, as this information is readily available in previous studies; instead, we concentrate on the essential factors radiology departments must consider when choosing AI solutions. These factors include clinical relevance, performance and validation, implementation and integration, clinical usability, costs and return on investment, and regulations, security, and privacy. We illustrate each factor with hypothetical scenarios to provide a clearer understanding and practical relevance. Through our experience and literature review, we provide insights and a practical roadmap for radiologists to navigate the complex landscape of AI in radiology. We aim to assist in making informed decisions that enhance diagnostic precision, improve patient outcomes, and streamline workflows, thus contributing to the advancement of radiological practices and patient care.

Evaluation of testicular tumour calcification with digital orchiography.

OBJECTIVES: To analyse the calcification of testicular tumours in the orchiectomy specimens detected by digital orchiography obtained in a full-field digital mammography (FFDM) unit. METHODS: Orchiectomy specimens of 37 consecutive patients were imaged by FFDM. Detected foci of calcification were stratified as: type 1, dense microcalcification; type 2, faint microcalcification; type 3, macrocalcification. Histopathology identified the tumour types, the presence of intratubular germ cell neoplasia (IGCN) and associated calcifications. Orchiography results correlated with the histopathology. RESULTS: On orchiography, 32/37of the specimens (86 %) had co-existing foci of calcification. Histopathology results revealed foci of calcification in 23/37 (62 %) of orchiectomy specimens. Of the 20 IGCN cases, 80 % presented with calcifications on orchiography. Fifty-six percent (14/25) of type 1, 70 % (12/17) of type 2, and 30 % (2/6) of type 3 foci of calcification were observed in IGCN-positive cases. CONCLUSION: This study classifies the morphology of testicular tumour calcification in three main groups by digital orchiography. In half of the testicular cancers, histopathologically proven IGCN is also found in addition to the index tumour. Type 2 foci of microcalcification detected by orchiography may be related to IGCN and may prompt further clinical assessment. KEY POINTS: • Orchiography can detect and classify calcification in 86 % of testicular cancers. • Intratubular germ cell neoplasia (IGCN) co-exists in 54 % of testicular cancers. • Type 2 foci of microcalcification detected by orchiography may be related to IGCN. • Orchiography may play a possible future role in the diagnosis of testicular IGCN.

Anatomically guided self-adapting deep neural network for clinically significant prostate cancer detection on bi-parametric MRI: a multi-center study.

OBJECTIVE: To evaluate the effectiveness of a self-adapting deep network, trained on large-scale bi-parametric MRI data, in detecting clinically significant prostate cancer (csPCa) in external multi-center data from men of diverse demographics; to investigate the advantages of transfer learning. METHODS: We used two samples: (i) Publicly available multi-center and multi-vendor Prostate Imaging: Cancer AI (PI-CAI) training data, consisting of 1500 bi-parametric MRI scans, along with its unseen validation and testing samples; (ii) In-house multi-center testing and transfer learning data, comprising 1036 and 200 bi-parametric MRI scans. We trained a self-adapting 3D nnU-Net model using probabilistic prostate masks on the PI-CAI data and evaluated its performance on the hidden validation and testing samples and the in-house data with and without transfer learning. We used the area under the receiver operating characteristic (AUROC) curve to evaluate patient-level performance in detecting csPCa. RESULTS: The PI-CAI training data had 425 scans with csPCa, while the in-house testing and fine-tuning data had 288 and 50 scans with csPCa, respectively. The nnU-Net model achieved an AUROC of 0.888 and 0.889 on the hidden validation and testing data. The model performed with an AUROC of 0.886 on the in-house testing data, with a slight decrease in performance to 0.870 using transfer learning. CONCLUSIONS: The state-of-the-art deep learning method using prostate masks trained on large-scale bi-parametric MRI data provides high performance in detecting csPCa in internal and external testing data with different characteristics, demonstrating the robustness and generalizability of deep learning within and across datasets. CLINICAL RELEVANCE STATEMENT: A self-adapting deep network, utilizing prostate masks and trained on large-scale bi-parametric MRI data, is effective in accurately detecting clinically significant prostate cancer across diverse datasets, highlighting the potential of deep learning methods for improving prostate cancer detection in clinical practice.