Search

Identifying optimal machine learning pipelines for computer-aided diagnosis is key for the development of robust, reproducible, and clinically relevant imaging biomarkers for endometrial carcinoma. The purpose of this study was to introduce the mathematical development of image descriptors computed from spherical harmonics (SPHARM) decompositions as well as the associated machine learning pipeline, and to evaluate their performance in predicting deep myometrial invasion (MI) and histopathological high-grade in preoperative multiparametric magnetic resonance imaging (MRI).This retrospective study included 128 women with histopathology-confirmed endometrial carcinomas who underwent 1.5-T MRI before hysterectomy between January 2011 and July 2015. SPHARM descriptors of each tumor were computed on multiparametric MRI images (T2-weighted, diffusion-weighted, dynamic contrast-enhanced-MRI and apparent diffusion coefficient maps). Tensor-based logistic regression was used to classify two-dimensional SPHARM rotationally-invariant descriptors. Head-to-head comparisons with radiomics analyses were performed with DeLong tests with Bonferroni-Holm correction to compare diagnostic performances.With all MRI contrasts, SPHARM analysis resulted in area under the curve, sensitivity, specificity, and balanced accuracy values of 0.94 (95% confidence interval [CI]: 0.85, 1.00), 100% (95% CI: 100, 100), 74% (95% CI: 51, 92), 87% (95% CI: 78, 98), respectively, for predicting deep MI. For predicting high-grade tumor histology, the corresponding values for the same diagnostic metrics were 0.81 (95% CI: 0.64, 0.90), 93% (95% CI: 67, 100), 63% (95% CI: 45, 79) and 78% (95% CI: 64, 86). The corresponding values achieved via radiomics were 0.92 (95% CI: 0.82, 0.95), 82% (95% CI: 65, 93), 80% (95% CI: 51, 94), 81% (95% CI: 70, 91) for deep MI and 0.72 (95% CI: 0.58, 0.83), 93% (95% CI: 65, 100), 55% (95% CI: 41, 69), 74% (95% CI: 52, 88) for high-grade histology. The diagnostic performance of the SPHARM analysis was not significantly different (P = 0.62) from that of radiomics for predicting deep MI but was significantly higher (P = 0.044) for predicting high-grade histology.The proposed SPHARM analysis yields similar or higher diagnostic performance than radiomics in identifying deep MI and high-grade status in histology-proven endometrial carcinoma.

To compare the performances of artificial intelligence (AI) to those of radiologists in wrist fracture detection on radiographs.This retrospective study included 637 patients (1917 radiographs) with wrist trauma between January 2017 and December 2019. The AI software used was a deep neuronal network algorithm. Ground truth was established by three senior musculoskeletal radiologists who compared the initial radiology reports (IRR) made by non-specialized radiologists, the results of AI, and the combination of AI and IRR (IR+AI) RESULTS: A total of 318 fractures were reported by the senior radiologists in 247 patients. Sensitivity of AI (83%; 95% CI: 78-87%) was significantly greater than that of IRR (76%; 95% CI: 70-81%) (p0.001). Specificities were similar for AI (96%; 95% CI: 93-97%) and for IRR (96%; 95% CI: 94-98%) (p = 0.80). The combination of AI+IRR had a significantly greater sensitivity (88%; 95% CI: 84-92%) compared to AI and IRR (p0.001) and a lower specificity (92%; 95% CI: 89-95%) (p0.001). The sensitivity for scaphoid fracture detection was acceptable for AI (84%) and IRR (80%) but poor for the detection of other carpal bones fracture (41% for AI and 26% for IRR).Performance of AI in wrist fracture detection on radiographs is better than that of non-specialized radiologists. The combination of AI and radiologist's analysis yields best performances.• Artificial intelligence has better performances for wrist fracture detection compared to non-expert radiologists in daily practice. • Performance of artificial intelligence greatly differs depending on the anatomical area. • Sensitivity of artificial intelligence for the detection of carpal bones fractures is 56%.

To develop a CT-based algorithm and evaluate its performance for the diagnosis of blunt bowel and/or mesenteric injury (BBMI) in patients with blunt abdominal trauma.This retrospective study included a training cohort of 79 patients (29 with BBMI and 50 patients with blunt abdominal trauma without BBMI) and a validation cohort of 37 patients (13 patients with BBMI and 24 patients with blunt abdominal trauma without BBMI). CT examinations were blindly analyzed by two independent radiologists. For each CT sign, the kappa value, sensitivity, specificity, and accuracy were calculated. A diagnostic algorithm was built using a recursive partitioning model on the training cohort, and its performances were assessed on the validation cohort.CT signs with kappa value0.6 were extraluminal gas, hemoperitoneum, no or moderate bowel wall enhancement, and solid organ injury. CT signs yielding best accuracies in the training cohort were extraluminal gas (98%; 95% CI: 91-100), bowel wall defect (97%; 95% CI: 91-100), irregularity of mesenteric vessels (97%; 95% CI: 90-99), and mesenteric vessel extravasation (97%; 95% CI: 90-99). Using a recursive partitioning model, a decision tree algorithm including extraluminal gas and no/moderate bowel wall enhancement was built, achieving 86% sensitivity (95% CI: 74-99) and 96% specificity (95% CI: 91-100) in the training cohort and 92% sensitivity (95% CI: 78-97) and 88% specificity (95% CI: 74-100) in the validation cohort for the diagnosis of BBMI.An effective diagnostic algorithm was built to identify BBMI in patients with blunt abdominal trauma using only extraluminal gas and no/moderate bowel wall enhancement on CT examination.• A CT diagnostic algorithm that included extraluminal gas and no/moderate bowel wall enhancement was built for the diagnosis of surgical blunt bowel and/or mesenteric injury. • A decision tree combining only two reproducible CT signs has high diagnostic performance for the diagnosis of surgical blunt bowel and/or mesenteric injury.

Pancreatic ductal carcinoma (PDAC) is one of the leading causes of cancer-related death worldwide. Computed tomography (CT) remains the primary imaging modality for diagnosis of PDAC. However, CT has limitations for early pancreatic tumor detection and tumor characterization so that it is currently challenged by magnetic resonance imaging. More recently, a particular attention has been given to radiomics for the characterization of pancreatic lesions using extraction and analysis of quantitative imaging features. In addition, radiomics has currently many applications that are developed in conjunction with artificial intelligence (AI) with the aim of better characterizing pancreatic lesions and providing a more precise assessment of tumor burden. This review article sums up recent advances in imaging of PDAC in the field of image/data acquisition, tumor detection, tumor characterization, treatment response evaluation, and preoperative planning. In addition, current applications of radiomics and AI in the field of PDAC are discussed.

The purpose of this study was to determine whether texture analysis features present on pretreatment unenhanced computed tomography (CT) images, derived from 18F-fluorodeoxyglucose positron emission/computed tomography (18-FDG PET/CT), can predict progression-free survival (PFS), progression-free survival at 24 months (PFS 24), time to next treatment (TTNT), and overall survival in patients with high-tumor-burden follicular lymphoma treated with immunochemotherapy and rituximab maintenance. Seventy-two patients with follicular lymphoma were retrospectively included. Texture analysis was performed on unenhanced CT images extracted from 18-FDG PET/CT examinations that were obtained within one month before treatment. Skewness at a fine texture scale (SSF = 2) was an independent predictor of PFS (hazard ratio = 3.72 (95% CI: 1.15, 12.11), p = 0.028), PFS 24 (hazard ratio = 13.38; 95% CI: 1.29, 138.13; p = 0.029), and TTNT (hazard ratio = 5.11; 95% CI: 1.18, 22.13; p = 0.029). Skewness values above −0.015 at SSF = 2 were significantly associated with lower PFS, PFS 24, and TTNT. Kurtosis without filtration was an independent predictor of PFS (SSF = 0; HR = 1.22 (95% CI: 1.04, 1.44), p = 0.013), and TTNT (SSF = 0; hazard ratio = 1.23; 95% CI: 1.04, 1.46; p = 0.013). This study shows that pretreatment unenhanced CT texture analysis-derived tumor skewness and kurtosis may be used as predictive biomarkers of PFS and TTNT in patients with high-tumor-burden follicular lymphoma treated with immunochemotherapy and rituximab maintenance.

PURPOSE The purpose of this study was to identify variables associated with extravasation on computed tomography angiography (CTA) in patients with hematochezia/melena and compare the outcome of patients with extravasation on CTA to those without extravasation. MATERIAL AND METHODS Ninety-four patients (51 men, 38 women; mean age, 69 ± 16 [SD] years) who underwent CTA within 30 days of hematochezia/melena were included. Variables associated with extravasation on CTA were searched using univariable and multivariable analyses. Outcomes of patients with visible extravasation on CTA were compared with those without visible extravasation. RESULTS One hundred and one CTA examinations were included. Extravasation was observed on 26/101 CTA examinations (26%). At multivariable analysis the need for vasopressor drugs (odds ratio [OR], 7.6; P = 0.040), high transfusion requirements (> 2 blood units) (OR, 7.1; P = 0.014), CTA performed on the day of a hemorrhagic event (OR, 46.2; P = 0.005) and repeat CTA (OR, 27.8; P = 0.011) were independently associated with extravasation on CTA. Extravasation on CTA was followed by a therapeutic procedure in 25/26 CTAs (96%; 26 patients) compared to 13/75 CTAs (17%; 68 patients) on which no extravasation was present (P < 0.001). No patients (0/26; 0%) with contrast extravasation on CTA died while 8 patients (8/61; 13%) without contrast extravasation died, although the difference was not significant (P = 0.099). CONCLUSION Extravasation on CTA in the setting of hematochezia or melena is especially seen in clinically unstable patients who receive more than two blood units. Presence of active extravasation on CTA leads to more frequent application of a therapeutic procedure; however, this does not significantly affect patient outcome.

Purpose The purpose of this study was to determine whether texture analysis features on pretreatment contrast-enhanced computed tomography (CT) images and their evolution can predict treatment response of metastatic skin melanoma (SM) treated with anti-PD1 monoclonal antibodies. Materials and methods Sixty patients (29 men, 31 women; median age, 56 years; age range: 27–91 years) with metastatic SM treated with pembrolizumab (43/60; 72%) or nivolumab (17/60; 28%) were included. Texture analysis of SM metastases was performed on baseline and first post-treatment evaluation CT examinations. Mean gray-level, entropy, kurtosis, skewness, and standard deviation values were derived from the pixel distribution histogram before and after spatial filtration at different anatomic scales, ranging from fine to coarse. Lasso penalized Cox regression analyses were performed to identify independent variables associated with favorable response to treatment. Results A total of 127 metastases were analyzed, with a median of two metastases per patient. Skewness at fine texture scale (spatial scale filtration [SSF] = 2; Hazard ratio [HR]: 3.51; 95% CI: 2.08–8.57; P = 0.010), skewness at medium texture scale (SSF = 3; HR: 0.56; 95% CI: 0.11–1.59; P = 0.014), variation of entropy at fine texture scale (SSF = 2; HR: 37.76; 95% CI: 3.48–496.22; P = 0.008) and LDH above the threshold of 248 UI/L (HR: 3.56; 95% CI: 1.78–21.35; P = 0.032] were independent predictors of response to treatment. Conclusion Pretreatment CT texture analysis-derived tumor skewness and variation of entropy between baseline and first control CT examination may be used as predictors of favorable response to anti-PD1 monoclonal antibodies in patients with metastatic SM.

To determine the potential of magnetic resonance-enterography (MRE) in the assessment of the anastomotic status in patients with Crohn disease and prior ileocolic resection.A total of 62 MRE examinations obtained in 52 patients with Crohn disease who had previously undergone ileocolic resection were retrospectively reviewed by two readers in consensus. MRE features (anastomotic wall thickening, wall stratification, wall enhancement pattern and degree, DWI signal intensity, ADC values, lymph nodes, comb sign and complications) were compared to clinical, endoscopic and histological findings that served as standard of reference. Sensitivity, specificity and accuracy of MRE were calculated.At univariate analysis, anastomotic wall thickening, anastomotic wall stratification, segmental wall enhancement, moderate wall enhancement, early and mucosal enhancement, and moderate/marked hyperintensity on diffusion-weighed imaging (DWI) were the most discriminative MRE features for differentiating between normal and abnormal anastomoses (p 0.001 for all variables). Anastomotic wall thickening and segmental anastomotic wall enhancement were the two most sensitive and accurate MRE variables for the diagnosis of abnormal anastomosis with sensitivities of 82% (95% CI: 67-92%) and accuracies of 84% (95% CI: 72-92%). At univariate analysis, hyperintensity on DWI of the anastomotic site was the most sensitive finding for distinguishing between inflammatory recurrence and fibrostenosis (sensitivity, 89%; 95% CI: 67-99%).MRE provides objective and relatively specific morphological criteria that help detect abnormal ileocolic anastomosis, but performances are lower when differentiating between inflammatory recurrence and fibrostenosis. DWI may be useful in identifying pathologic anastomosis and, in particular, in distinguishing between inflammatory recurrence and fibrostenosis.

Purpose The purpose of this study was to report the prevalence and imaging features of abdominal metastases from Merkel cell carcinoma (MCC) on computed tomography (CT) examinations. Materials and methods A total of 111 patients with MCC from two institutions were initially identified. Of these, 27 patients (27/111; 24.3%) had abdominal metastases from MCC present on CT examination. There were 19 men and 8 women with a mean age of 75 ± 10.8 (SD) years (age range: 46–92 years). CT examinations were retrospectively reviewed by two radiologists and analyzed quantitatively for the number and dimensions of abdominal metastases from MCC and qualitatively in terms of location, margins, contours, homogeneity, patterns of enhancement, vascular involvement and extension of metastases from MCC. Results Fifteen patients (15/27; 56%) had abdominal metastatic disease at initial diagnosis and twelve (12/27; 44%) developed abdominal metastases during the course of the disease. The mean number of locations of abdominal metastases was 2.1 ± 1.12 (SD) (range: 1–5). Abdominal metastases involved abdominal lymph nodes (16/27; 59%), adrenal glands/kidneys/retroperitoneum (14/27; 52%), mesentery/peritoneum (13/27; 48%), liver (7/27; 26%) and pancreas (7/27; 26%). Vascular involvement was found in association with peritoneal/mesenteric metastases in 6/13 (46%) patients with intraperitoneal metastases or in association with abdominal lymph nodes in 4/16 (25%) patients. Ureteral encasement and/or dilatation was found in 4/14 (28%) patients with retroperitoneal metastases and 3/16 (19%) patients with abdominal lymph nodes. Metastases to the liver, pancreas, peritoneum, retroperitoneum and adrenal glands displayed internal enhancement during the arterial phase in 1/2 (50%), 4/5 (80%), 4/7 (57%) and 5/8 (62%) patients for whom arterial phase was available, respectively. Conclusion : Metastases from MCC have a prevalence of 24.3% on CT examination and may involve a variety of abdominal organs, mainly lymph nodes, peritoneal and retroperitoneal structures, but also the liver and pancreas. CT features of abdominal metastases from MCC include hypervascularity during the arterial phase of enhancement and eventually vascular and ureteral involvement.

Sarcopenia, defined as decreased muscle mass and strength, can be evaluated by a computed tomography (CT) examination and might be associated with reduced survival in patients with carcinoma. The prognosis of patients with metastatic pancreatic carcinoma is poor. The FOLFIRINOX (a combination of 5-fluorouracil, irinotecan, and oxaliplatin) chemotherapy regimen is a validated first-line treatment option. We investigated the impact of sarcopenia on overall survival (OS) and progression-free survival (PFS) in patients with metastatic pancreatic carcinoma. Clinical data and CT examinations of patients treated with FOLFIRINOX were retrospectively reviewed. Sarcopenia was estimated using baseline CT examinations. Seventy-five patients were included. Forty-three (57.3%) were classified as sarcopenic. The median OS of non-sarcopenic and sarcopenic patients were 15.6 and 14.1 months, respectively (p = 0.36). The median PFS was 10.3 in non-sarcopenic patients and 9.3 in sarcopenic patients (p = 0.83). No differences in toxicity of FOLFIRINOX were observed. There was a trend towards a higher probability of short-term death (within 4 months of diagnosis) in sarcopenic patients. In this study, the detection of sarcopenia failed to predict a longer OS or PFS in selected patients deemed eligible by a physician for triplet chemotherapy and receiving the FOLFIRINOX regimen in a first-line setting, confirming the major importance of a comprehensive patient assessment by physicians in selecting the best treatment option.

International audience; The emergence of massively parallel yet affordable computing devices has been a game changer for research in the field of artificial intelligence (AI). In addition, dramatic investment from the web giants has fostered the development of a high-quality software stack. Going forward, the combination of faster computers with dedicated software libraries and the widespread availability of data has opened the door to more flexibility in the design of AI models. Radiomics is a process used to discover new imaging biomarkers that has multiple applications in radiology and can be used in conjunction with AI. AI can be used throughout the various processes of diagnostic imaging, including data acquisition, reconstruction, analysis and reporting. Today, the concept of "AI-augmented" radiologists is preferred to the theory of the replacement of radiologists by AI in many indications. Current evidence bolsters the assumption that AI-assisted radiologists work better and faster. Interventional radiology becomes a data-rich specialty where the entire procedure is fully recorded in a standardized DICOM format and accessible via standard picture archiving and communication systems. No other interventional specialty can bolster such readiness. In this setting, interventional radiology could lead the development of AI-powered applications in the broader interventional community. This article provides an update on the current status of radiomics and AI research, analyzes upcoming challenges and also discusses the main applications in AI in interventional radiology to help radiologists better understand and criticize articles reporting AI in medical imaging.

Background: Despite its toxicity, modified FOLFIRINOX is the main chemotherapy for localized, operable pancreatic adenocarcinomas. Sarcopenia is known as a factor in lower overall survival (OS). The purpose of this study was to assess the impact of sarcopenia on OS in patients with localized pancreatic ductal adenocarcinoma (PDAC) who received modified FOLFIRINOX or gemcitabine as adjuvant chemotherapy. Methods: Patients with operated PDAC who received gemcitabine-based (GEM group) or oxaliplatin-based (OXA group) adjuvant chemotherapy between 2008 and 2021 were retrospectively included. Sarcopenia was estimated on a baseline computed tomography (CT) examination using the skeletal muscular index (SMI). The primary evaluation criterion was OS. Secondary evaluation criteria were disease-free survival (DFS) and toxicity. Results: Seventy patients treated with gemcitabine-based (n = 49) and oxaliplatin-based (n = 21) chemotherapy were included, with a total of fifteen sarcopenic patients (eight in the GEM group and seven in the OXA group). The median OS was shorter in sarcopenic patients (25 months) compared to non-sarcopenic patients (158 months) (p = 0.01). A longer OS was observed in GEM non-sarcopenic patients (158 months) compared to OXA sarcopenic patients (14.4 months) (p < 0.01). The median OS was 157.7 months in the GEM group vs. 34.1 months in the OXA group (p = 0.13). No differences in median DFS were found between the GEM group and OXA group. More toxicity events were observed in the OXA group (50%) than in the GEM group (10%), including vomiting (p = 0.02), mucositis (p = 0.01) and neuropathy (p = 0.01). Conclusion: Sarcopenia is associated with a worse prognosis in patients with localized operated PDAC whatever the delivered adjuvant chemotherapy.

The spleen can be affected by a variety of diseases. Some of them are readily identified as variations of normal or benign diseases on imaging. However, for a substantial number of focal splenic abnormalities, the diagnosis can be difficult so that histopathologic analysis may be required for a definite diagnosis. In this review, the typical splenic abnormalities that can be diagnosed with imaging with a high degree of confidence are illustrated. The complementary role of computed tomography (CT), magnetic resonance imaging and positron emission tomography/CT that helps make a diagnostic approach is discussed. Finally, current applications and future trends of radiomics and artificial intelligence for the diagnosis of splenic diseases are addressed.

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer and currently the third-leading cause of cancer-related death worldwide. Recently, artificial intelligence (AI) has emerged as an important tool to improve clinical management of HCC, including for diagnosis, prognostication and evaluation of treatment response. Different AI approaches, such as machine learning and deep learning, are both based on the concept of developing prediction algorithms from large amounts of data, or big data. The era of digital medicine has led to a rapidly expanding amount of routinely collected health data which can be leveraged for the development of AI models. Various studies have constructed AI models by using features extracted from ultrasound imaging, computed tomography imaging and magnetic resonance imaging. Most of these models have used convolutional neural networks. These tools have shown promising results for HCC detection, characterization of liver lesions and liver/tumor segmentation. Regarding treatment, studies have outlined a role for AI in evaluation of treatment response and improvement of pre-treatment planning. Several challenges remain to fully integrate AI models in clinical practice. Future research is still needed to robustly evaluate AI algorithms in prospective trials, and improve interpretability, generalizability and transparency. If such challenges can be overcome, AI has the potential to profoundly change the management of patients with HCC. The purpose of this review was to sum up current evidence on AI approaches using imaging for the clinical management of HCC.

International audience

In the elective field of adrenal imaging, artificial intelligence (AI) can be used for adrenal lesion detection, characterization, hypersecreting syndrome management and patient follow-up. Although a perfect AI tool that includes all required steps from detection to analysis does not exist yet, multiple AI algorithms have been developed and tested with encouraging results. However, AI in this setting is still at an early stage. In this regard, most published studies about AI in adrenal gland imaging report preliminary results that do not have yet daily applications in clinical practice. In this review, recent developments and current results of AI in the field of adrenal imaging are presented. Limitations and future perspectives of AI are discussed.

Background Stratifying high-risk histopathologic features in endometrial carcinoma is important for treatment planning. Radiomics analysis at preoperative MRI holds potential to identify high-risk phenotypes. Purpose To evaluate the performance of multiparametric MRI three-dimensional radiomics-based machine learning models for differentiating low- from high-risk histopathologic markers-deep myometrial invasion (MI), lymphovascular space invasion (LVSI), and high-grade status-and advanced-stage endometrial carcinoma. Materials and Methods This dual-center retrospective study included women with histologically proven endometrial carcinoma who underwent 1.5-T MRI before hysterectomy between January 2011 and July 2015. Exclusion criteria were tumor diameter less than 1 cm, missing MRI sequences or histopathology reports, neoadjuvant therapy, and malignant neoplasms other than endometrial carcinoma. Three-dimensional radiomics features were extracted after tumor segmentation at MRI (T2-weighted, diffusion-weighted, and dynamic contrast-enhanced MRI). Predictive features were selected in the training set with use of random forest (RF) models for each end point, and trained RF models were applied to the external test set. Five board-certified radiologists conducted MRI-based staging and deep MI assessment in the training set. Areas under the receiver operating characteristic curve (AUCs) were reported with balanced accuracies, and radiologists' readings were compared with radiomics with use of McNemar tests. Results In total, 157 women were included: 94 at the first institution (training set; mean age, 66 years ± 11 [SD]) and 63 at the second institution (test set; 67 years ± 12). RF models dichotomizing deep MI, LVSI, high grade, and International Federation of Gynecology and Obstetrics (FIGO) stage led to AUCs of 0.81 (95% CI: 0.68, 0.88), 0.80 (95% CI: 0.67, 0.93), 0.74 (95% CI: 0.61, 0.86), and 0.84 (95% CI: 0.72, 0.92), respectively, in the test set. In the training set, radiomics provided increased performance compared with radiologists' readings for identifying deep MI (balanced accuracy, 86% vs 79%; P = .03), while no evidence of a difference was observed in performance for advanced FIGO stage (80% vs 78%; P = .27). Conclusion Three-dimensional radiomics can stratify patients by using preoperative MRI according to high-risk histopathologic end points in endometrial carcinoma and provide nonsignificantly different or higher performance than radiologists in identifying advanced stage and deep myometrial invasion, respectively. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Kido and Nishio in this issue.

The purpose of this study was to report the effectiveness of preoperative transcatheter arterial embolization (TAE) of musculoskeletal tumors in terms of blood loss and functional outcomes. Patients who underwent preoperative TAE of hypervascular musculoskeletal tumors between January 2018 and December 2021 were retrospectively included. The patients’ characteristics, TAE procedure details, degree of post-TAE devascularization, surgical outcomes in terms of red blood cell transfusion and functional results were collected. The degree of devascularization was compared between patients who had peri-operative transfusion and those who did not. Thirty-one patients were included. The 31 TAE procedures led to complete (58%) or near-complete (42%) tumor devascularization. Twenty-two patients (71%) had no blood transfusion during surgery. Nine patients (29%) had a blood transfusion, with a median number of red blood cell packs of three (q1, 2; q3, 4; range: 1–4). Eight patients (27%) had complete improvement of the initial musculoskeletal symptoms at the end of the follow-up, 15 (50%) had partially satisfying improvement, 4 (13%) had partially unsatisfying improvement and 3 (10%) had no improvement. Our study suggests that preoperative TAE of hypervascular musculoskeletal tumors allowed for bloodless surgery in 71% of patients and minimal transfusion needs for the remaining 29%.

The objective of this study was to assess the performance of ultrasound and magnetic resonance imaging (MRI) features in helping to classify the type of placenta accreta spectrum (PAS; accreta/increta vs percreta), alone or combined in a predictive score.We conducted a retrospective study in 82 pregnant women with PAS who underwent ultrasound and MRI examination of the pelvis before delivery (from an initial cohort of 185 women with PAS). We estimated the sensitivity, specificity and accuracy of MRI and ultrasound in the diagnosis of the type of PAS. We analyzed cesarean and imaging features using univariable logistic regression analysis. We constructed a nomogram to predict the risk of placenta percreta and validated it with bootstrap resampling, then used receiver operating characteristic curves to assess the performance of the model in distinguishing between placenta percreta and placenta accreta/increta.Among the 82 patients, 29 (35%) had placenta accreta/increta and 53 (65%) had placenta percreta. The best features to discriminate between placenta accreta/increta and placenta percreta with ultrasound were increased vascularization at the uterine serosa-bladder wall interface (odds ratio [OR] 7.93; 95% confidence interval [CI] 2.78-24.99; p 0.01) and the number of lacunae without a hyperechogenic halo (OR 1.36; 95% CI 1.14-1.67; p = 0.012). Concerning MRI markers, heterogeneous placenta (OR 12.89; 95% CI 3.05-89.16; p = 0.002), dark intraplacental bands (OR 12.89; 95% CI 3.05-89.16; p = 0.002) and bladder wall interruption (OR 15.89; 95% CI 4.78-73.33; p 0.001) had a higher OR in discriminating placenta accreta/increta from placenta percreta. The nomogram yielded areas under the curve of 0.841 (95% CI 0.754-0.927) and 0.856 (95% CI 0.767-0.945), after bootstrap resampling, for the accurate prediction of placenta percreta.The nomogram we developed to predict the risk of placenta percreta among patients with PAS had good discriminative capabilities. This performance and its impact on maternal morbidity should be confirmed by future prospective studies.

Hepatocellular carcinoma (HCC) usually occurs in the setting of liver cirrhosis and more rarely in a healthy liver. Its incidence has increased in the past years, especially in western countries with the rising prevalence of non-alcoholic fatty liver disease. The prognosis of advanced HCC is low. In the first-line setting of advanced HCC, sorafenib, a tyrosine kinase inhibitor, was the only validated treatment for many years. In 2020, the combination of atezolizumab, an immune checkpoint inhibitor, and bevacizumab showed superiority to sorafenib alone in survival, making it the first-line recommended treatment. Regorafenib and lenvatinib, other multikinase inhibitors, were also validated in the second and first-line settings, respectively. Transarterial chemoembolization can be an alternative treatment for patients with intermediate-stage HCC and preserved liver function, including unresectable multinodular HCC without extrahepatic spread. The current challenge in advanced HCC lies in the selection of a patient for the optimal treatment, taking into account the underlying liver disease and liver function. Indeed, all trial patients present with a Child–Pugh score of A, and the optimal approach for other patients is still unclear. Furthermore, the combination of atezolizumab and bevacizumab should be considered in the absence of medical contraindication. Many trials testing immune checkpoint inhibitors in association with anti-angiogenic agents are ongoing, and primary results are promising. The landscape in advanced HCC management is undergoing profound change, and many challenges remain for optimal patient management in the years to come. This review aimed to provide an overview of current systemic treatment options for patients with advanced unresectable HCC who are not candidates for liver-directed therapy.

Radiomics is a relatively new approach for image analysis. As a part of radiomics, texture analysis, which consists in extracting a great amount of quantitative data from original images, can be used to identify specific features that can help determining the actual nature of a pancreatic lesion and providing other information such as resectability, tumor grade, tumor response to neoadjuvant therapy or survival after surgery. In this review, the basic of radiomics, recent developments and the results of texture analysis using computed tomography and magnetic resonance imaging in the field of pancreatic tumors are presented. Future applications of radiomics, such as artificial intelligence, are discussed.

Esophageal peroral endoscopic myotomy (POEM) is the treatment of reference of major obstructive esophageal motility disorders but the detection of early complications remains challenging. Our aim was to report the radiological findings on meglumine diatrizoate esophagograms after esophageal POEM and identify variables associated with patient outcomes. The imaging and clinical files of 106 patients who underwent POEM for achalasia or other major obstructive esophageal motility disorders were retrospectively analyzed. Post POEM esophagograms were reviewed for the presence of pneumoperitoneum, pleural effusion, extraesophageal contrast leakage, and dislocated clips. Associations between length of hospital stay and radiological findings were searched for using a Cox multivariate analysis. A total of 106 patients (M/F = 56/50; mean age = 50 ± 2 [SD] years) underwent 106 POEM procedures with a meglumine diatrizoate esophagogram on postoperative day 1. Overall median hospital stay was 3 days (range 1–20 days). Pneumoperitoneum, pleural effusion, extraesophageal contrast leakage, and dislocated clips were observed in 90/106 (84.9%), 12/106 (11.3%), 4/106 (3.8%), and 0/106 (0%) patients, respectively. At multivariate analysis, pleural effusion (p = 0.005; adjusted hazard ratio [aHR] = 0.35 [95% CI 0.17–0.73]) and extraesophageal contrast leakage (p = 0.039; aHR = 0.27 [95% CI 0.08–0.94]) were associated with a prolonged hospital stay. Pneumoperitoneum was not associated with unfavorable outcome (p = 0.99). Pneumoperitoneum is a common finding after POEM and is not indicative of unfavorable patient outcome. Conversely, post POEM pleural effusion and extraesophageal contrast leakage are associated with a longer hospital stay. • Water-soluble esophagogram is a valid diagnostic modality to diagnose early complications after esophageal endoscopic myotomy for esophageal motility disorders. • At multivariate analysis, pleural effusion and extraesophageal contrast leakage are associated with a prolonged hospital stay after peroral endoscopic myotomy. • Pneumoperitoneum is not associated with unfavorable outcome after peroral endoscopic myotomy.

This study was conducted in order to establish the joint Society of Abdominal Radiology (SAR) and European Society of Urogenital Radiology (ESUR) guidelines on placenta accreta spectrum (PAS) disorders and propose strategies to standardize image acquisition, interpretation, and reporting for this condition with MRI. The published evidence-based data and the opinion of experts were combined using the RAND–UCLA Appropriateness Method and formed the basis for these consensus guidelines. The responses of the experts to questions regarding the details of patient preparation, MRI protocol, image interpretation, and reporting were collected, analyzed, and classified as “recommended” versus “not recommended” (if at least 80% consensus among experts) or uncertain (if less than 80% consensus among experts). Consensus regarding image acquisition, interpretation, and reporting was determined using the RAND–UCLA Appropriateness Method. The use of a tailored MRI protocol and standardized report was recommended. A standardized imaging protocol and reporting system ensures recognition of the salient features of PAS disorders. These consensus recommendations should be used as a guide for the evaluation of PAS disorders with MRI. • MRI is a powerful adjunct to ultrasound and provides valuable information on the topography and depth of placental invasion. • Consensus statement proposed a common lexicon to allow for uniformity in MRI acquisition, interpretation, and reporting of PAS disorders. • Seven MRI features, namely intraplacental dark T2 bands, uterine/placental bulge, loss of low T2 retroplacental line, myometrial thinning/disruption, bladder wall interruption, focal exophytic placental mass, and abnormal vasculature of the placental bed, reached consensus and are categorized as “recommended” for diagnosing PAS disorders.

Perihilar cholangiocarcinoma (PHC) is a common and highly intractable malignancy of the main biliary tree confluence. PHC is associated with a poor prognosis because of its insidious local spread that makes it challenging to diagnose and assess. Surgical resection remains the standard curative treatment (up to 50% 5-year overall survival after negative-margin resection). More aggressive surgical approaches have recently emerged, pushing the boundaries of PHC resectability at the cost of a higher morbidity. As such, adequate preoperative preparation (i.e., biliary drainage, venous embolization) is now regarded as a critical issue to increase the number of patients amenable to extended liver resection. Thorough imaging plays a pivotal role in the preoperative setting in both PHC resectability assessment and patient preparation to surgery. Despite recent improvement in PHC imaging, its assessment remains challenging and only 50-60% of patients who are scheduled to undergo surgery are ultimately amenable to curative resection. Therefore, a knowledge of available diagnostic and interventional imaging techniques is important to improve PHC management. Herein, we review the various imaging techniques and preoperative radiological interventions such as biliary drainage, portal vein embolization and liver venous deprivation that are available in PHC management focusing on the anatomical and oncological considerations that are crucial to prepare and guide curative surgical resection.

Simple Summary As we are facing an increasing incidence of pancreatic neuroendocrine tumors (pNETs), it appears necessary to better classify this disease—and in 2017 the WHO classification introduced a new category of well differentiated grade 3 tumors. pNETs treatment requires some specific background and recent reviews on the multimodal management of this disease are missing. Indeed, those modalities constantly evolve and this review, focusing on Grade 1 and Grade 2 pNETs, aims to clarify both oncological and surgical treatment options from localized tumors to multi metastatic disease. Every aspect of the strategy are discussed in this review from the oncologist and the surgeon’s point of view and with a special focus on a minimally invasive approach. Abstract Pancreatic neuroendocrine tumors (p-NETs) are rare tumors with a recent growing incidence. In the 2017 WHO classification, p-NETs are classified into well-differentiated (i.e., p-NETs grade 1 to 3) and poorly differentiated neuroendocrine carcinomas (i.e., p-NECs). P-NETs G1 and G2 are often non-functioning tumors, of which the prognosis depends on the metastatic status. In the localized setting, p-NETs should be surgically managed, as no benefit for adjuvant chemotherapy has been demonstrated. Parenchymal sparing resection, including both duodenum and pancreas, are safe procedures in selected patients with reduced endocrine and exocrine long-term dysfunction. When the p-NET is benign or borderline malignant, this surgical option is associated with low rates of severe postoperative morbidity and in-hospital mortality. This narrative review offers comments, tips, and tricks from reviewing the available literature on these different options in order to clarify their indications. We also sum up the overall current data on p-NETs G1 and G2 management.

Objectives: To investigate intra-specialty citation patterns of radiology articles, compared with another medical specialty: gastroenterology/hepatology. Methods: Four radiology journals ( Radiology, European Radiology, Diagnostic and Interventional Imaging, Canadian Association of Radiologists Journal) and four gastroenterology/hepatology journals ( Journal of Hepatology, Journal of Gastroenterology, World Journal of Gastroenterology, Journal of Clinical Gastroenterology) with similar Web of Science in-category 2020 IF ranking were selected. The original research, review, letter, and editorial articles published in these journals in 2021 were identified. The average number of intra-specialty citations per article (intra-specialty citation count) and percentage of intra-specialty citations out of total citations per article (intra-specialty citation rate) were compared between radiology and gastroenterology/hepatology articles using Student’s t-test. Results: The radiology articles demonstrated a lower total citation count per article (radiology: 29.7 ± .4 (mean ± SEM), n = 2063; gastroenterology/hepatology: 50.1 ± 1.4, n = 1335). The intra-specialty citation count was also lower in radiology articles than gastroenterology/hepatology articles (radiology: 12.9 ± .2, gastroenterology/hepatology: 19.6 ± .7; P < .001), both overall and in all article types. Additionally, the overall intra-specialty citation rate was not significantly different between the two specialties (radiology: 48.8% ± .5%; gastroenterology/hepatology: 47.1 ± .8%; P = .057), although the intra-specialty citation rates were higher in radiology original research and editorial article types. Conclusions: The significantly lower per-article intra-specialty citation counts in all radiology article types, a measurement that directly links to specialty IFs, may contribute to the lower impact factors of radiology journals compared with gastroenterology/hepatology ones.

To prospectively determine the value of post-MRI micro-ultrasonography (microUS) in the diagnosis of transition zone (TZ) significant prostate cancer (sPCa).Eighty-four consecutive men (66 ± 6.3 years) with a mean PSA level of 10.2 ± 7.4 ng/mL and at least one TZ-PI-RADS2 lesion were included. All patients had MRI-directed microUS and biopsy. Sensitivity and specificity of post-MRI microUS to visualize PI-RADS2 TZ lesions, the cancer detection rate of TZ-sPCa, and tumor characteristics according to their visibility on microUS were evaluated. Interreader agreement for detecting microUS+ lesions was evaluated using Cohen's kappa test.Of the 92 PI-RADS2 lesions, 71 (71/92; 77%) were visible on microUS and biopsy was performed without image fusion, which was required for the 21 invisible lesions (21/92; 22.8%). TZ-sPCa detection rate was 51.1% (47/92). Sensitivity and specificity of MRI-directed microUS were 83% (39/47; 95% CI: 69.2-92.4%) and 28.9% (13/45; 95% CI: 16.4-44.3%), on a per-lesion basis and 86.4% (38/45; 95% CI: 72.6-94.8%) and 27.5% (11/40; 95% CI: 14.6-43.9%) on a per-patient basis. Visible tumors on microUS exhibited a larger volume and a lower mean ADC value than non-visible tumors (15.8 ± 5.1 vs. 12.5 ± 3.6 mm and 0.82 ± 1.1 × 10MRI-targeted transrectal microUS is effective to detect TZ-sPCa. TRUS-MRI image fusion helps overcome limitations due to TZ tissue heterogeneity.microUS can visualize the majority of MRI-detected PI-RADS2 TZ lesions (sensitivity = 83%). Interreader agreement of MRI-directed microUS in the detection of TZ lesions appears excellent (kappa = 0.88). In 77% of PI-RADS2 TZ lesions, biopsy was performed under microUS visual control. MRI fusion system was only used to overcome limitations due to tissue heterogeneity of benign prostatic hyperplasia.

To assess whether heterogeneous adrenal adenomas can be distinguished from heterogeneous non-adenomas with Computed Tomography (CT) and/or Magnetic Resonance Imaging (MRI).From 2009 to 2019, 980 consecutive adrenalectomies were retrospectively identified. Patients without adequate CT/MRI, with homogeneous and/or 1 cm lesions were excluded. Differences between adenomas and non-adenomas were analyzed using Chi-square, Student t or Fischer tests, and interobserver agreement using weighted kappa test or intraclass correlation coefficient. Independent variables associated with adenomas were searched for using multivariable analysis. Area under the receiver operating characteristic curve (AUC) of the final model and its diagnostic performances were calculated.Final population comprised 183 patients (106 women, 77 men, mean age 53.2 ± 14.4 years) with 124 non-adenomas and 59 heterogeneous adenomas. Macroscopic or microscopic fat on CT/MRI allowed diagnosis of adenoma with 98% specificity and 63% sensitivity. Interobserver agreement was almost perfect for macroscopic fat (k = 0.82; 95% CI 0.66; 0.94) and substantial for microscopic fat (k = 0.75; 95% CI 0.62; 0.86). A multivariable model including micro- or macroscopic fat [Odds ratio (OR) 81.19; 95% CI 20.17; 572.27], diameter 5.5 cm (OR 7.32; 95% CI 2.17; 31.28), calcifications (OR 5.68; 95% CI 2.08; 16.18), and hemorrhage (OR 3.10; 95% CI 0.70; 15.35) had an AUC of 0.91 (95% CI 0.86; 0.96), 71% (42/59, 95% CI 58; 82) sensitivity, 93% (115/124; 95% CI 87; 97) specificity, and 86% (157/183; 95% CI 79; 90) accuracy for the diagnosis of adenoma.A multivariable model enables CT/MR diagnosis of heterogeneous adenomas. Presence of microscopic fat, even if partial, in a heterogeneous mass is highly specific of adenoma.