Combining Deep Learning and Handcrafted Radiomics for Classification of Suspicious Lesions on Contrast-enhanced Mammograms.

Background Handcrafted radiomics and deep learning (DL) models individually achieve good performance in lesion classification (benign vs malignant) on contrast-enhanced mammography (CEM) images. Purpose To develop a comprehensive machine learning tool able to fully automatically identify, segment, and classify breast lesions on the basis of CEM images in recall patients. Materials and Methods CEM images and clinical data were retrospectively collected between 2013 and 2018 for 1601 recall patients at Maastricht UMC+ and 283 patients at Gustave Roussy Institute for external validation. Lesions with a known status (malignant or benign) were delineated by a research assistant overseen by an expert breast radiologist. Preprocessed low-energy and recombined images were used to train a DL model for automatic lesion identification, segmentation, and classification. A handcrafted radiomics model was also trained to classify both human- and DL-segmented lesions. Sensitivity for identification and the area under the receiver operating characteristic curve (AUC) for classification were compared between individual and combined models at the image and patient levels. Results After the exclusion of patients without suspicious lesions, the total number of patients included in the training, test, and validation data sets were 850 (mean age, 63 years ± 8 [SD]), 212 (62 years ± 8), and 279 (55 years ± 12), respectively. In the external data set, lesion identification sensitivity was 90% and 99% at the image and patient level, respectively, and the mean Dice coefficient was 0.71 and 0.80 at the image and patient level, respectively. Using manual segmentations, the combined DL and handcrafted radiomics classification model achieved the highest AUC (0.88 [95% CI: 0.86, 0.91]) (P < .05 except compared with DL, handcrafted radiomics, and clinical features model, where P = .90). Using DL-generated segmentations, the combined DL and handcrafted radiomics model showed the highest AUC (0.95 [95% CI: 0.94, 0.96]) (P < .05). Conclusion The DL model accurately identified and delineated suspicious lesions on CEM images, and the combined output of the DL and handcrafted radiomics models achieved good diagnostic performance. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Bahl and Do in this issue.

Novel Functional Features of cGMP Substrate Proteins IRAG1 and IRAG2.

The inositol triphosphate-associated proteins IRAG1 and IRAG2 are cGMP kinase substrate proteins that regulate intracellular Ca2+. Previously, IRAG1 was discovered as a 125 kDa membrane protein at the endoplasmic reticulum, which is associated with the intracellular Ca2+ channel IP3R-I and the PKGIß and inhibits IP3R-I upon PKGIß-mediated phosphorylation. IRAG2 is a 75 kDa membrane protein homolog of IRAG1 and was recently also determined as a PKGI substrate. Several (patho-)physiological functions of IRAG1 and IRAG2 were meanwhile elucidated in a variety of human and murine tissues, e.g., of IRAG1 in various smooth muscles, heart, platelets, and other blood cells, of IRAG2 in the pancreas, heart, platelets, and taste cells. Hence, lack of IRAG1 or IRAG2 leads to diverse phenotypes in these organs, e.g., smooth muscle and platelet disorders or secretory deficiency, respectively. This review aims to highlight the recent research regarding these two regulatory proteins to envision their molecular and (patho-)physiological tasks and to unravel their functional interplay as possible (patho-)physiological counterparts.

IRAG2 Interacts with IP3-Receptor Types 1, 2, and 3 and Regulates Intracellular Ca2+ in Murine Pancreatic Acinar Cells.

The inositol 1,4,5-triphosphate receptor-associated 2 (IRAG2) is also known as Jaw1 or lymphoid-restricted membrane protein (LRMP) and shares homology with the inositol 1,4,5-triphosphate receptor-associated cGMP kinase substrate 1 (IRAG1). IRAG1 interacts with inositol trisphosphate receptors (IP3 receptors /IP3R) via its coiled-coil domain and modulates Ca2+ release from intracellular stores. Due to the homology of IRAG1 and IRAG2, especially in its coiled-coil domain, it is possible that IRAG2 has similar interaction partners like IRAG1 and that IRAG2 also modulates intracellular Ca2+ signaling. In our study, we localized IRAG2 in pancreatic acinar cells of the exocrine pancreas, and we investigated the interaction of IRAG2 with IP3 receptors and its impact on intracellular Ca2+ signaling and exocrine pancreatic function, like amylase secretion. We detected the interaction of IRAG2 with different subtypes of IP3R and altered Ca2+ release in pancreatic acinar cells from mice lacking IRAG2. IRAG2 deficiency decreased basal levels of intracellular Ca2+, suggesting that IRAG2 leads to activation of IP3R under unstimulated basal conditions. Moreover, we observed that loss of IRAG2 impacts the secretion of amylase. Our data, therefore, suggest that IRAG2 modulates intracellular Ca2+ signaling, which regulates exocrine pancreatic function.

[Intercontinental Multidisciplinary Oncology Videoconferencing between the South Pacific and the French mainland: Results after one year and 323 cases of rare or complex cancers discussed]. / RCP Pacifique : bilan après un an et 323 cancers rares et/ou complexes discutés.

INTRODUCTION: Overseas France represents 18 % of French territory and is home to 4 % of its population for whom there is unequal treatment in the field of rare/complex cancer. AIM: To report our experience of intercontinental multidisciplinary videoconferencing between the French mainland and Pacific territories. METHODS: Every other friday, three centers located in Papeete, Nouméa and Paris-Villejuif connected between 6:30 AM and 8:00 AM GMT to discuss cases of rare/complex cancers. RESULTS: Between November 2019 and December 2020, 323 presentations implicating 233 patients involved sarcoma (n=93), digestive pathology (n=60), neuroendocrine tumors (n=35), urology (n=24), gynecology (n=24), neurology (n=16), thyroid pathology (n=14), dermatology (n=14), senology (n=11), hematology (n=11), ENT pathology (n=10), pathology thoracic (n=10) and pediatrics (n=1). Of the 233 patients, 134 (57.5 %) living in New Caledonia and 99 (42.5 %) in French Polynesia, 117 (50.5 %) had metastatic disease. 39 patients (16.7 %) were transferred to French mainland (EVASAN), for surgery (n=25), vectorized radiotherapy (n=7), biopsy (n=5), chemotherapy (n=1) or inclusion in a clinical trial (n=1). 195 patients (83.7 %) were treated at home, 15 (6.4 %) are still awaiting a decision and 4 (1.7 %) lost to follow-up. CONCLUSION: The use of videoconferencing to discuss rare/complex cancer cases was effective in guaranteeing French overseas population access to innovative therapies and clinical trials, limiting the need for intercontinental transfer to 16.7 %.

Loss of PKGIß/IRAG1 Signaling Causes Anemia-Associated Splenomegaly.

Inositol 1,4,5-triphosphate receptor-associated cGMP kinase substrate 1 (IRAG1) is a substrate protein of the NO/cGMP-signaling pathway and forms a ternary complex with the cGMP-dependent protein kinase Iß (PKGIß) and the inositol triphosphate receptor I (IP3R-I). Functional studies about IRAG1 exhibited that IRAG1 is specifically phosphorylated by the PKGIß, regulating cGMP-mediated IP3-dependent Ca2+-release. IRAG1 is widely distributed in murine tissues, e.g., in large amounts in smooth muscle-containing tissues and platelets, but also in lower amounts, e.g., in the spleen. The NO/cGMP/PKGI signaling pathway is important in several organ systems. A loss of PKGI causes gastrointestinal disorders, anemia and splenomegaly. Due to the similar tissue distribution of the PKGIß to IRAG1, we investigated the pathophysiological functions of IRAG1 in this context. Global IRAG1-KO mice developed gastrointestinal bleeding, anemia-associated splenomegaly and iron deficiency. Additionally, Irag1-deficiency altered the protein levels of some cGMP/PKGI signaling proteins-particularly a strong decrease in the PKGIß-in the colon, spleen and stomach but did not change mRNA-expression of the corresponding genes. The present work showed that a loss of IRAG1 and the PKGIß/IRAG1 signaling has a crucial function in the development of gastrointestinal disorders and anemia-associated splenomegaly. Furthermore, global Irag1-deficient mice are possible in vivo model to investigate PKGIß protein functions.

Intercontinental Multidisciplinary Oncology Videoconferencing for Rare and Complex Cancer: An Alternative to Systematic Transfer.

PURPOSE: To report our experience of intercontinental multidisciplinary oncology videoconferencing between the French mainland and South Pacific to discuss rare and/or complex cancer cases. METHODS: On the first and third Friday of each month, all participants connected between 6:30 am and 8:00 am GMT to discuss using a web conference service. RESULTS: Between November 2019 and April 2020, 99 cases concerning 78 patients were discussed. Oncology subspecialties required were sarcoma (n = 36), digestive (n = 29), dermatology (n = 5), gynecology (n = 5), breast (n = 5), urology (n = 5), hematology (n = 5), ENT (n = 3), thoracic (n = 3), thyroid (n = 2), and pediatric (n = 1). Median patient age was 58 years, 41 were female (53%), 37 were male (47%), and 43 had a metastatic disease (55%). Following discussion, 16 patients (21%) were transferred to the French mainland. Reasons for transfer were requirement for complex surgery (n = 11) and need for specialized diagnostic biopsy (n = 5). Fifty-six patients were treated locally, with systemic chemotherapy (n = 36), surveillance (n = 8), surgery (n = 8), radiotherapy (n = 3), or endoscopy (n = 1). Direct benefits for patients treated in their local facility included strategy changes (surveillance or surgery contraindication, n = 9), targeted therapy decision (n = 14), immunotherapy decision (n = 9), and diagnostic or metastatic status corrections (n = 4). Six patients are still awaiting decision. CONCLUSION: Using real-time intercontinental multidisciplinary oncology videoconferencing to discuss complex or rare cancer cases is reliable and effective for decision making. This concept helped to limit to 21% the need for transfers to the mainland.

Comprehensive Gene Expression Analysis Detects Global Reduction of Proteasome Subunits in Schizophrenia.

BACKGROUND: The main challenge in the study of schizophrenia is its high heterogeneity. While it is generally accepted that there exist several biological mechanisms that may define distinct schizophrenia subtypes, they have not been identified yet. We performed comprehensive gene expression analysis to search for molecular signals that differentiate schizophrenia patients from healthy controls and examined whether an identified signal was concentrated in a subgroup of the patients. METHODS: Transcriptome sequencing of 14 superior temporal gyrus (STG) samples of subjects with schizophrenia and 15 matched controls from the Stanley Medical Research Institute (SMRI) was performed. Differential expression and pathway enrichment analysis results were compared to an independent cohort. Replicability was tested on 6 additional independent datasets. RESULTS: The 2 STG cohorts showed high replicability. Pathway enrichment analysis of the down-regulated genes pointed to proteasome-related pathways. Meta-analysis of differential expression identified down-regulation of 12 of 39 proteasome subunit genes in schizophrenia. The signal of proteasome subunits down-regulation was replicated in 6 additional datasets (overall 8 cohorts with 267 schizophrenia and 266 control samples, from 5 brain regions). The signal was concentrated in a subgroup of patients with schizophrenia. CONCLUSIONS: We detected global down-regulation of proteasome subunits in a subgroup of patients with schizophrenia. We hypothesize that the down-regulation of proteasome subunits leads to proteasome dysfunction that causes accumulation of ubiquitinated proteins, which has been recently detected in a subgroup of schizophrenia patients. Thus, down-regulation of proteasome subunits might define a biological subtype of schizophrenia.

IRAG1 Deficient Mice Develop PKG1ß Dependent Pulmonary Hypertension.

PKGs are serine/threonine kinases. PKG1 has two isoforms-PKG1α and ß. Inositol trisphosphate receptor (IP3R)-associated cGMP-kinase substrate 1 (IRAG1) is a substrate for PKG1ß. IRAG1 is also known to further interact with IP3RI, which mediates intracellular Ca2+ release. However, the role of IRAG1 in PH is not known. Herein, WT and IRAG1 KO mice were kept under normoxic or hypoxic (10% O2) conditions for five weeks. Animals were evaluated for echocardiographic variables and went through right heart catheterization. Animals were further sacrificed to prepare lungs and right ventricular (RV) for immunostaining, western blotting, and pulmonary artery smooth muscle cell (PASMC) isolation. IRAG1 is expressed in PASMCs and downregulated under hypoxic conditions. Genetic deletion of IRAG1 leads to RV hypertrophy, increase in RV systolic pressure, and RV dysfunction in mice. Absence of IRAG1 in lung and RV have direct impacts on PKG1ß expression. Attenuated PKG1ß expression in IRAG1 KO mice further dysregulates other downstream candidates of PKG1ß in RV. IRAG1 KO mice develop PH spontaneously. Our results indicate that PKG1ß signaling via IRAG1 is essential for the homeostasis of PASMCs and RV. Disturbing this signaling complex by deleting IRAG1 can lead to RV dysfunction and development of PH in mice.

Incorporating Magnetic Resonance Imaging (MRI) Based Radiation Therapy Response Prediction into Clinical Practice for Locally Advanced Cervical Cancer Patients.

In recent years, magnetic resonance imaging (MRI) has become one of the standard imaging tools to define the macroscopic gross tumor volume in locally advanced cervical cancer patients based on T2-weighted sequence. Recent data suggest that functional MRI could be used to potentially improve the delineation of target volumes based on physiologic features, defining radioresistant subvolumes that may require higher doses to achieve local cure. Functional imaging can be used to predict tumor biology and outcome, as well as for assessment of tumor response during radiotherapy. The concept of adaptive radiotherapy relies on the possibility of monitoring variations in target volumes structures to guide treatment-plan modification during radiotherapy, taking into account not only internal movements but also tumor response. With integrated MRI in radiotherapy linear accelerators, motion monitoring during treatment delivery has become available. MRI can be also used to accurately evaluate cervical tumor residual volume after chemoradiotherapy, and therefore allowing a personalized treatment planning for brachytherapy boost, based on tumor radiosensitivity. In this review, we discuss how MRI tumor response assessment could be included into clinical practice during radiation therapy in locally advanced cervical cancer patients.