Deep-learning-based reconstruction of undersampled MRI to reduce scan times: a multicentre, retrospective, cohort study.

BACKGROUND: The extended acquisition times required for MRI limit its availability in resource-constrained settings. Consequently, accelerating MRI by undersampling k-space data, which is necessary to reconstruct an image, has been a long-standing but important challenge. We aimed to develop a deep convolutional neural network (dCNN) optimisation method for MRI reconstruction and to reduce scan times and evaluate its effect on image quality and accuracy of oncological imaging biomarkers. METHODS: In this multicentre, retrospective, cohort study, MRI data from patients with glioblastoma treated at Heidelberg University Hospital (775 patients and 775 examinations) and from the phase 2 CORE trial (260 patients, 1083 examinations, and 58 institutions) and the phase 3 CENTRIC trial (505 patients, 3147 examinations, and 139 institutions) were used to develop, train, and test dCNN for reconstructing MRI from highly undersampled single-coil k-space data with various acceleration rates (R=2, 4, 6, 8, 10, and 15). Independent testing was performed with MRIs from the phase 2/3 EORTC-26101 trial (528 patients with glioblastoma, 1974 examinations, and 32 institutions). The similarity between undersampled dCNN-reconstructed and original MRIs was quantified with various image quality metrics, including structural similarity index measure (SSIM) and the accuracy of undersampled dCNN-reconstructed MRI on downstream radiological assessment of imaging biomarkers in oncology (automated artificial intelligence-based quantification of tumour burden and treatment response) was performed in the EORTC-26101 test dataset. The public NYU Langone Health fastMRI brain test dataset (558 patients and 558 examinations) was used to validate the generalisability and robustness of the dCNN for reconstructing MRIs from available multi-coil (parallel imaging) k-space data. FINDINGS: In the EORTC-26101 test dataset, the median SSIM of undersampled dCNN-reconstructed MRI ranged from 0·88 to 0·99 across different acceleration rates, with 0·92 (95% CI 0·92-0·93) for 10-times acceleration (R=10). The 10-times undersampled dCNN-reconstructed MRI yielded excellent agreement with original MRI when assessing volumes of contrast-enhancing tumour (median DICE for spatial agreement of 0·89 [95% CI 0·88 to 0·89]; median volume difference of 0·01 cm3 [95% CI 0·00 to 0·03] equalling 0·21%; p=0·0036 for equivalence) or non-enhancing tumour or oedema (median DICE of 0·94 [95% CI 0·94 to 0·95]; median volume difference of -0·79 cm3 [95% CI -0·87 to -0·72] equalling -1·77%; p=0·023 for equivalence) in the EORTC-26101 test dataset. Automated volumetric tumour response assessment in the EORTC-26101 test dataset yielded an identical median time to progression of 4·27 months (95% CI 4·14 to 4·57) when using 10-times-undersampled dCNN-reconstructed or original MRI (log-rank p=0·80) and agreement in the time to progression in 374 (95·2%) of 393 patients with data. The dCNN generalised well to the fastMRI brain dataset, with significant improvements in the median SSIM when using multi-coil compared with single-coil k-space data (p<0·0001). INTERPRETATION: Deep-learning-based reconstruction of undersampled MRI allows for a substantial reduction of scan times, with a 10-times acceleration demonstrating excellent image quality while preserving the accuracy of derived imaging biomarkers for the assessment of oncological treatment response. Our developments are available as open source software and hold considerable promise for increasing the accessibility to MRI, pending further prospective validation. FUNDING: Deutsche Forschungsgemeinschaft (German Research Foundation) and an Else Kröner Clinician Scientist Endowed Professorship by the Else Kröner Fresenius Foundation.

Artificial intelligence and radiologists in prostate cancer detection on MRI (PI-CAI): an international, paired, non-inferiority, confirmatory study.

BACKGROUND: Artificial intelligence (AI) systems can potentially aid the diagnostic pathway of prostate cancer by alleviating the increasing workload, preventing overdiagnosis, and reducing the dependence on experienced radiologists. We aimed to investigate the performance of AI systems at detecting clinically significant prostate cancer on MRI in comparison with radiologists using the Prostate Imaging-Reporting and Data System version 2.1 (PI-RADS 2.1) and the standard of care in multidisciplinary routine practice at scale. METHODS: In this international, paired, non-inferiority, confirmatory study, we trained and externally validated an AI system (developed within an international consortium) for detecting Gleason grade group 2 or greater cancers using a retrospective cohort of 10 207 MRI examinations from 9129 patients. Of these examinations, 9207 cases from three centres (11 sites) based in the Netherlands were used for training and tuning, and 1000 cases from four centres (12 sites) based in the Netherlands and Norway were used for testing. In parallel, we facilitated a multireader, multicase observer study with 62 radiologists (45 centres in 20 countries; median 7 [IQR 5-10] years of experience in reading prostate MRI) using PI-RADS (2.1) on 400 paired MRI examinations from the testing cohort. Primary endpoints were the sensitivity, specificity, and the area under the receiver operating characteristic curve (AUROC) of the AI system in comparison with that of all readers using PI-RADS (2.1) and in comparison with that of the historical radiology readings made during multidisciplinary routine practice (ie, the standard of care with the aid of patient history and peer consultation). Histopathology and at least 3 years (median 5 [IQR 4-6] years) of follow-up were used to establish the reference standard. The statistical analysis plan was prespecified with a primary hypothesis of non-inferiority (considering a margin of 0·05) and a secondary hypothesis of superiority towards the AI system, if non-inferiority was confirmed. This study was registered at ClinicalTrials.gov, NCT05489341. FINDINGS: Of the 10 207 examinations included from Jan 1, 2012, through Dec 31, 2021, 2440 cases had histologically confirmed Gleason grade group 2 or greater prostate cancer. In the subset of 400 testing cases in which the AI system was compared with the radiologists participating in the reader study, the AI system showed a statistically superior and non-inferior AUROC of 0·91 (95% CI 0·87-0·94; p<0·0001), in comparison to the pool of 62 radiologists with an AUROC of 0·86 (0·83-0·89), with a lower boundary of the two-sided 95% Wald CI for the difference in AUROC of 0·02. At the mean PI-RADS 3 or greater operating point of all readers, the AI system detected 6·8% more cases with Gleason grade group 2 or greater cancers at the same specificity (57·7%, 95% CI 51·6-63·3), or 50·4% fewer false-positive results and 20·0% fewer cases with Gleason grade group 1 cancers at the same sensitivity (89·4%, 95% CI 85·3-92·9). In all 1000 testing cases where the AI system was compared with the radiology readings made during multidisciplinary practice, non-inferiority was not confirmed, as the AI system showed lower specificity (68·9% [95% CI 65·3-72·4] vs 69·0% [65·5-72·5]) at the same sensitivity (96·1%, 94·0-98·2) as the PI-RADS 3 or greater operating point. The lower boundary of the two-sided 95% Wald CI for the difference in specificity (-0·04) was greater than the non-inferiority margin (-0·05) and a p value below the significance threshold was reached (p<0·001). INTERPRETATION: An AI system was superior to radiologists using PI-RADS (2.1), on average, at detecting clinically significant prostate cancer and comparable to the standard of care. Such a system shows the potential to be a supportive tool within a primary diagnostic setting, with several associated benefits for patients and radiologists. Prospective validation is needed to test clinical applicability of this system. FUNDING: Health~Holland and EU Horizon 2020.

Onset of spin entanglement in doped carbon nanotubes studied by EPR.

Nanoscale semiconductors with isolated spin impurities have been touted as promising materials for their potential use at the intersection of quantum, spin, and information technologies. Electron paramagnetic resonance (EPR) studies of spins in semiconducting carbon nanotubes have overwhelmingly focused on spins more strongly localized by sp3-type lattice defects. However, the creation of such impurities is irreversible and requires specific reactions to generate them. Shallow charge impurities, on the other hand, are more readily and widely produced by simple redox chemistry, but have not yet been investigated for their spin properties. Here, we use EPR to study p-doped (6,5) semiconducting single-wall carbon nanotubes (s-SWNTs) and elucidate the role of impurity-impurity interactions in conjunction with exchange and correlation effects for the spin behavior of this material. A quantitative comparison of the EPR signals with phenomenological modeling combined with configuration interaction electronic structure calculations of impurity pairs shows that orbital overlap, combined with exchange and correlation effects, causes the EPR signal to disappear due to spin entanglement for doping levels corresponding to impurity spacings of 14 nm (at 30 K). This transition is predicted to shift to higher doping levels with increasing temperature and to lower levels with increasing screening, providing an opportunity for improved spin control in doped s-SWNTs.

Special Issue: Molecular Ecology, Physiology and Biochemistry of Insects 3.0.

With the outstanding work of Sir Vincent B [...].

Villification of the intestinal epithelium is driven by Foxl1.

The primitive gut tube of mammals initially forms as a simple cylinder consisting of the endoderm-derived, pseudostratified epithelium and the mesoderm-derived surrounding mesenchyme. During mid-gestation a dramatic transformation occurs in which the epithelium is both restructured into its final cuboidal form and simultaneously folded and refolded to create intestinal villi and intervillus regions, the incipient crypts. Here we show that the mesenchymal winged helix transcription factor Foxl1, itself induced by epithelial hedgehog signaling, controls villification by activating BMP and PDGFRα as well as planar cell polarity genes in epithelial-adjacent telocyte progenitors, both directly and in a feed-forward loop with Foxo3.

Separation of telomere protection from length regulation by two different point mutations at amino acid 492 of RTEL1.

RTEL1 is an essential DNA helicase that plays multiple roles in genome stability and telomere length regulation. A variant of RTEL1 with a lysine at position 492 is associated with short telomeres in Mus spretus , while a conserved methionine at this position is found in M. musculus, which has ultra-long telomeres. In humans, a missense mutation at this position ( RTEL1 M492I ) causes a fatal telomere biology disease termed Hoyeraal-Hreidarsson syndrome (HHS). We previously described a M. musculus mouse model termed 'Telomouse', in which changing methionine 492 to a lysine (M492K) shortened the telomeres to their length in humans. Here, we report on the derivation of a mouse strain carrying the M492I mutation, termed 'HHS mouse'. The HHS mouse telomeres are not as short as those of Telomice but nevertheless they display higher levels of telomeric DNA damage, fragility and recombination, associated with anaphase bridges and micronuclei. These observations indicate that the two mutations separate critical functions of RTEL1: M492K mainly reduces the telomere length setpoint, while M492I predominantly disrupts telomere protection. The two mouse models enable dissecting the mechanistic roles of RTEL1 and the different contributions of short telomeres and DNA damage to telomere biology diseases, genomic instability, cancer, and aging.

Telocyte Recruitment During the Emergence of a Metaplastic Niche in the Stomach.

BACKGROUND & AIM: Telocytes, a recently identified type of subepithelial interstitial cell, have garnered attention for their potential roles in tissue homeostasis and repair. However, their contribution to gastric metaplasia remains unexplored. This study elucidates the role of telocytes in the development of metaplasia within the gastric environment. METHODS: To investigate the presence and behavior of telocytes during metaplastic transitions, we used drug-induced acute injury models (using DMP-777 or L635) and a genetically engineered mouse model (Mist1-Kras). Lineage tracing via the Foxl1-CreERT2;R26R-tdTomato mouse model was used to track telocyte migratory dynamics. Immunofluorescence staining was used to identify telocyte markers and evaluate their correlation with metaplasia-related changes. RESULTS: We confirmed the existence of FOXL1+/PDGFRα+ double-positive telocytes in the stomach's isthmus region. As metaplasia developed, we observed a marked increase in the telocyte population. The distribution of telocytes expanded beyond the isthmus to encompass the entire gland and closely reflected the expansion of the proliferative cell zone. Rather than a general response to mucosal damage, the shift in telocyte distribution was associated with the establishment of a metaplastic cell niche at the gland base. Furthermore, lineage-tracing experiments highlighted the active recruitment of telocytes to the emerging metaplastic cell niche, and we observed expression of Wnt5a, Bmp4, and Bmp7 in PDGFRα+ telocytes. CONCLUSIONS: These results suggest that telocytes contribute to the evolution of a gastric metaplasia niche. The dynamic behavior of these stromal cells, their responsiveness to metaplastic changes, and potential association with Wnt5a, Bmp4, and Bmp7 signaling emphasize the significance of telocytes in tissue adaptation and repair.

Modeling type 1 diabetes progression using machine learning and single-cell transcriptomic measurements in human islets.

Type 1 diabetes (T1D) is a chronic condition in which beta cells are destroyed by immune cells. Despite progress in immunotherapies that could delay T1D onset, early detection of autoimmunity remains challenging. Here, we evaluate the utility of machine learning for early prediction of T1D using single-cell analysis of islets. Using gradient-boosting algorithms, we model changes in gene expression of single cells from pancreatic tissues in T1D and non-diabetic organ donors. We assess if mathematical modeling could predict the likelihood of T1D development in non-diabetic autoantibody-positive donors. While most autoantibody-positive donors are predicted to be non-diabetic, select donors with unique gene signatures are classified as T1D. Our strategy also reveals a shared gene signature in distinct T1D-associated models across cell types, suggesting a common effect of the disease on transcriptional outputs of these cells. Our study establishes a precedent for using machine learning in early detection of T1D.

No evidence for persistent enteroviral B infection of pancreatic islets in type 1 diabetic and pre-diabetic patients from RNA-Seq data.

Persistent enterovirus B infection has been proposed as an important contributor to the etiology of type 1 diabetes. We leveraged extensive bulk RNA-sequencing data from alpha, beta, and exocrine cells, as well as single cell islet RNA-Seq data from the Human Pancreas Analysis Program (HPAP), to evaluate the presence of enterovirus B sequences in the pancreas of type 1 diabetic and pre-diabetic (non-diabetic but auto-antibody positive) patients. We examined all available HPAP data for either assay type, including non-diabetic, type 1, and type 2 diabetic donors. To assess the presence of viral reads, we analyzed all reads not mapping to the human genome with the taxonomic classification system Kraken2 and its full viral database, augmented to encompass representatives for all (28) enterovirus B serotypes for which a complete genome is available. As a secondary approach, we input the same sequence reads into the STAR aligner using these 28 enterovirus B genomes as the reference. No enterovirus B sequences were detected by either approach in any of the 243 bulk RNA libraries nor in any of the 79 single cell RNA libraries. While we cannot rule out the possibility of a very low-grade persistent enterovirus B infection in the donors analyzed, our data do not support the notion of chronic viral infection by these viruses as a major driver of type 1 diabetes.

Multiplexed Imaging Mass Cytometry Analysis Characterizes the Vascular Niche in Pancreatic Cancer.

Oncogenesis and progression of pancreatic ductal adenocarcinoma (PDAC) are driven by complex interactions between the neoplastic component and the tumor microenvironment, which includes immune, stromal, and parenchymal cells. In particular, most PDACs are characterized by a hypovascular and hypoxic environment that alters tumor cell behavior and limits the efficacy of chemotherapy and immunotherapy. Characterization of the spatial features of the vascular niche could advance our understanding of inter- and intratumoral heterogeneity in PDAC. In this study, we investigated the vascular microenvironment of PDAC by applying imaging mass cytometry using a 26-antibody panel on 35 regions of interest across 9 patients, capturing more than 140,000 single cells. The approach distinguished major cell types, including multiple populations of lymphoid and myeloid cells, endocrine cells, ductal cells, stromal cells, and endothelial cells. Evaluation of cellular neighborhoods identified 10 distinct spatial domains, including multiple immune and tumor-enriched environments as well as the vascular niche. Focused analysis revealed differential interactions between immune populations and the vasculature and identified distinct spatial domains wherein tumor cell proliferation occurs. Importantly, the vascular niche was closely associated with a population of CD44-expressing macrophages enriched for a proangiogenic gene signature. Taken together, this study provides insights into the spatial heterogeneity of PDAC and suggests a role for CD44-expressing macrophages in shaping the vascular niche. Significance: Imaging mass cytometry revealed that pancreatic ductal cancers are composed of 10 distinct cellular neighborhoods, including a vascular niche enriched for macrophages expressing high levels of CD44 and a proangiogenic gene signature.

AnnoSpat annotates cell types and quantifies cellular arrangements from spatial proteomics.

Cellular composition and anatomical organization influence normal and aberrant organ functions. Emerging spatial single-cell proteomic assays such as Image Mass Cytometry (IMC) and Co-Detection by Indexing (CODEX) have facilitated the study of cellular composition and organization by enabling high-throughput measurement of cells and their localization directly in intact tissues. However, annotation of cell types and quantification of their relative localization in tissues remain challenging. To address these unmet needs for atlas-scale datasets like Human Pancreas Analysis Program (HPAP), we develop AnnoSpat (Annotator and Spatial Pattern Finder) that uses neural network and point process algorithms to automatically identify cell types and quantify cell-cell proximity relationships. Our study of data from IMC and CODEX shows the higher performance of AnnoSpat in rapid and accurate annotation of cell types compared to alternative approaches. Moreover, the application of AnnoSpat to type 1 diabetic, non-diabetic autoantibody-positive, and non-diabetic organ donor cohorts recapitulates known islet pathobiology and shows differential dynamics of pancreatic polypeptide (PP) cell abundance and CD8+ T cells infiltration in islets during type 1 diabetes progression.

Identifying long-term survivors and those at higher or lower risk of relapse among patients with cytogenetically normal acute myeloid leukemia using a high-dimensional mixture cure model.

Patients with cytogenetically normal acute myeloid leukemia (CN-AML) may harbor prognostically relevant gene mutations and thus be categorized into one of the three 2022 European LeukemiaNet (ELN) genetic-risk groups. Nevertheless, there remains heterogeneity with respect to relapse-free survival (RFS) within these genetic-risk groups. Our training set included 306 adults on Alliance for Clinical Trials in Oncology studies with de novo CN-AML aged < 60 years who achieved a complete remission and for whom centrally reviewed cytogenetics, RNA-sequencing, and gene mutation data from diagnostic samples were available (Alliance trial A152010). To overcome deficiencies of the Cox proportional hazards model when long-term survivors are present, we developed a penalized semi-parametric mixture cure model (MCM) to predict RFS where RNA-sequencing data comprised the predictor space. To validate model performance, we employed an independent test set from the German Acute Myeloid Leukemia Cooperative Group (AMLCG) consisting of 40 de novo CN-AML patients aged < 60 years who achieved a complete remission and had RNA-sequencing of their pre-treatment sample. For the training set, there was a significant non-zero cure fraction (p = 0.019) with 28.5% of patients estimated to be cured. Our MCM included 112 genes associated with cure, or long-term RFS, and 87 genes associated with latency, or shorter-term time-to-relapse. The area under the curve and C-statistic were respectively, 0.947 and 0.783 for our training set and 0.837 and 0.718 for our test set. We identified a novel, prognostically relevant molecular signature in CN-AML, which allows identification of patient subgroups independent of 2022 ELN genetic-risk groups.Trial registration Data from companion studies CALGB 8461, 9665 and 20202 (trials registered at www.clinicaltrials.gov as, respectively, NCT00048958, NCT00899223, and NCT00900224) were obtained from Alliance for Clinical Trials in Oncology under data sharing study A152010. Data from the AMLCG 2008 trial was registered at www.clinicaltrials.gov as NCT01382147.

Clonal hematopoiesis and its impact on the aging osteo-hematopoietic niche.

Clonal hematopoiesis (CH) defines a premalignant state predominantly found in older persons that increases the risk of developing hematologic malignancies and age-related inflammatory diseases. However, the risk for malignant transformation or non-malignant disorders is variable and difficult to predict, and defining the clinical relevance of specific candidate driver mutations in individual carriers has proved to be challenging. In addition to the cell-intrinsic mechanisms, mutant cells rely on and alter cell-extrinsic factors from the bone marrow (BM) niche, which complicates the prediction of a mutant cell's fate in a shifting pre-malignant microenvironment. Therefore, identifying the insidious and potentially broad impact of driver mutations on supportive niches and immune function in CH aims to understand the subtle differences that enable driver mutations to yield different clinical outcomes. Here, we review the changes in the aging BM niche and the emerging evidence supporting the concept that CH can progressively alter components of the local BM microenvironment. These alterations may have profound implications for the functionality of the osteo-hematopoietic niche and overall bone health, consequently fostering a conducive environment for the continued development and progression of CH. We also provide an overview of the latest technology developments to study the spatiotemporal dependencies in the CH BM niche, ideally in the context of longitudinal studies following CH over time. Finally, we discuss aspects of CH carrier management in clinical practice, based on work from our group and others.

The evolutionarily ancient FOXA transcription factors shape the murine gut microbiome via control of epithelial glycosylation.

Evolutionary adaptation of multicellular organisms to a closed gut created an internal microbiome differing from that of the environment. Although the composition of the gut microbiome is impacted by diet and disease state, we hypothesized that vertebrates promote colonization by commensal bacteria through shaping of the apical surface of the intestinal epithelium. Here, we determine that the evolutionarily ancient FOXA transcription factors control the composition of the gut microbiome by establishing favorable glycosylation on the colonic epithelial surface. FOXA proteins bind to regulatory elements of a network of glycosylation enzymes, which become deregulated when Foxa1 and Foxa2 are deleted from the intestinal epithelium. As a direct consequence, microbial composition shifts dramatically, and spontaneous inflammatory bowel disease ensues. Microbiome dysbiosis was quickly reversed upon fecal transplant into wild-type mice, establishing a dominant role for the host epithelium, in part mediated by FOXA factors, in controlling symbiosis in the vertebrate holobiont.

DNA Methylation-Based Assessment of Cell Composition in Human Pancreas and Islets.

Assessment of pancreas cell type composition is crucial to the understanding of the genesis of diabetes. Current approaches use immunodetection of protein markers, for example, insulin as a marker of ß-cells. A major limitation of these methods is that protein content varies in physiological and pathological conditions, complicating the extrapolation to actual cell number. Here, we demonstrate the use of cell type-specific DNA methylation markers for determining the fraction of specific cell types in human islet and pancreas specimens. We identified genomic loci that are uniquely demethylated in specific pancreatic cell types and applied targeted PCR to assess the methylation status of these loci in tissue samples, enabling inference of cell type composition. In islet preparations, normalization of insulin secretion to ß-cell DNA revealed similar ß-cell function in pre-type 1 diabetes (T1D), T1D, and type 2 diabetes (T2D), which was significantly lower than in donors without diabetes. In histological pancreas specimens from recent-onset T1D, this assay showed ß-cell fraction within the normal range, suggesting a significant contribution of ß-cell dysfunction. In T2D pancreata, we observed increased α-cell fraction and normal ß-cell fraction. Methylation-based analysis provides an accurate molecular alternative to immune detection of cell types in the human pancreas, with utility in the interpretation of insulin secretion assays and the assessment of pancreas cell composition in health and disease.

Intra-islet α-cell Gs signaling promotes glucagon release.

Glucagon, a hormone released from pancreatic α-cells, is critical for maintaining euglycemia and plays a key role in the pathophysiology of diabetes. To stimulate the development of new classes of therapeutic agents targeting glucagon release, key α-cell signaling pathways that regulate glucagon secretion need to be identified. Here, we focused on the potential importance of α-cell Gs signaling on modulating α-cell function. Studies with α-cell-specific mouse models showed that activation of α-cell Gs signaling causes a marked increase in glucagon secretion. We also found that intra-islet adenosine plays an unexpected autocrine/paracrine role in promoting glucagon release via activation of α-cell Gs-coupled A2A adenosine receptors. Studies with α-cell-specific Gαs knockout mice showed that α-cell Gs also plays an essential role in stimulating the activity of the Gcg gene, thus ensuring proper islet glucagon content. Our data suggest that α-cell enriched Gs-coupled receptors represent potential targets for modulating α-cell function for therapeutic purposes.

Cellular dynamics following CAR T cell therapy are associated with response and toxicity in relapsed/refractory myeloma.

B-cell maturation antigen (BCMA)-targeting chimeric antigen receptor (CAR) T cells revolutionized the treatment of relapsed/refractory multiple myeloma (RRMM). However, data on cellular (CAR) T cell dynamics and the association with response, resistance or the occurrence of cytokine release syndrome (CRS) are limited. Therefore, we performed a comprehensive flow cytometry analysis of 27 RRMM patients treated with Idecabtagene vicleucel (Ide-cel) to assess the expansion capacity, persistence and effects on bystander cells of BCMA-targeting CAR T cells. Additionally, we addressed side effects, like cytokine release syndrome (CRS) and cytopenia. Our results show that in vivo expansion of CD8+ CAR T cells is correlated to response, however persistence is not essential for durable remission in RRMM patients. In addition, our data provide evidence, that an increased fraction of CD8+ T cells at day of leukapheresis in combination with successful lymphodepletion positively influence the outcome. We show that patients at risk for higher-grade CRS can be identified already prior to lymphodepletion. Our extensive characterization contributes to a better understanding of the dynamics and effects of BCMA-targeting CAR T cells, in order to predict the response of individual patients as well as side effects, which can be counteracted at an early stage or even prevented.

Fair evaluation of federated learning algorithms for automated breast density classification: The results of the 2022 ACR-NCI-NVIDIA federated learning challenge.

The correct interpretation of breast density is important in the assessment of breast cancer risk. AI has been shown capable of accurately predicting breast density, however, due to the differences in imaging characteristics across mammography systems, models built using data from one system do not generalize well to other systems. Though federated learning (FL) has emerged as a way to improve the generalizability of AI without the need to share data, the best way to preserve features from all training data during FL is an active area of research. To explore FL methodology, the breast density classification FL challenge was hosted in partnership with the American College of Radiology, Harvard Medical Schools' Mass General Brigham, University of Colorado, NVIDIA, and the National Institutes of Health National Cancer Institute. Challenge participants were able to submit docker containers capable of implementing FL on three simulated medical facilities, each containing a unique large mammography dataset. The breast density FL challenge ran from June 15 to September 5, 2022, attracting seven finalists from around the world. The winning FL submission reached a linear kappa score of 0.653 on the challenge test data and 0.413 on an external testing dataset, scoring comparably to a model trained on the same data in a central location.

Shared and unique 3D genomic features of substance use disorders across multiple cell types.

Recent genome-wide association studies (GWAS) have revealed shared genetic components among alcohol, opioid, tobacco and cannabis use disorders. However, the extent of the underlying shared causal variants and effector genes, along with their cellular context, remain unclear. We leveraged our existing 3D genomic datasets comprising high-resolution promoter-focused Capture-C/Hi-C, ATAC-seq and RNA-seq across >50 diverse human cell types to focus on genomic regions that coincide with GWAS loci. Using stratified LD regression, we determined the proportion of genomewide SNP heritability attributable to the features assayed across our cell types by integrating recent GWAS summary statistics for the relevant traits: alcohol use disorder (AUD), tobacco use disorder (TUD), opioid use disorder (OUD) and cannabis use disorder (CanUD). Statistically significant enrichments (P<0.05) were observed in 14 specific cell types, with heritability reaching 9.2-fold for iPSC-derived cortical neurons and neural progenitors, confirming that they are crucial cell types for further functional exploration. Additionally, several pancreatic cell types, notably pancreatic beta cells, showed enrichment for TUD, with heritability enrichments up to 4.8-fold, suggesting genomic overlap with metabolic processes. Further investigation revealed significant positive genetic correlations between T2D with both TUD and CanUD (FDR<0.05) and a significant negative genetic correlation with AUD. Interestingly, after partitioning the heritability for each cell type's cis-regulatory elements, the correlation between T2D and TUD for pancreatic beta cells was greater (r=0.2) than the global genetic correlation value. Our study provides new genomic insights into substance use disorders and implicates cell types where functional follow-up studies could reveal causal variant-gene mechanisms underpinning these disorders.