Association between adherence to posttreatment National Comprehensive Cancer Network (NCCN) surveillance guidelines and detection of recurrent uterine cancer.

OBJECTIVE: To identify correlations between disease recurrence and adherence to NCCN posttreatment surveillance guidelines in patients who develop recurrent uterine cancer. METHODS: Retrospective analysis identified patients (n = 60) with recurrent uterine cancer and at least one surveillance visit with a gynecologic oncologist between 2011 and 2020. Adherence to NCCN guidelines and details of recurrence were recorded. RESULTS: Recurrent uterine cancer was identified in 60 patients with an average time to recurrence (TTR) of 25 months. Of those, 39 (65%) were adherent to NCCN surveillance guidelines and 36 (60%) were symptomatic at the time of recurrence diagnosis. Asymptomatic recurrence was diagnosed by imaging in 11 (46%), physical exam in 7 (29%), and blood work in 6 (25%) patients. Patients who were adherent to NCCN guidelines were diagnosed with recurrence on average 11 months earlier (p = 0.0336). Adherence was an independent predictor of TTR for all patients regardless of symptoms. There was no significant effect of age, race, primary language, or stage of disease on adherence. CONCLUSION: Adherence to NCCN posttreatment surveillance guidelines for uterine cancer is independently associated with an earlier diagnosis of recurrence.

The association of black race with receipt of hysterectomy and survival in low-risk endometrial cancer.

OBJECTIVE: To determine whether Black race is associated with treatment and survival among women with low-risk endometrial cancer. METHODS: Black and White women with Stage IA grade 1-2 endometrioid endometrial carcinoma diagnosed from 2010 to 2016 in the SEER 18 dataset were identified (n = 23,431), and clinical and socioeconomic attributes obtained. Five-year cancer-specific survival (CSS) and relative survival (RS) were calculated using SEER*Stat 8.3.9. Cox proportional hazards model was used to determine predictors of overall survival (OS) and CSS. RESULTS: There was a significantly higher proportion of Black women who did not have surgery compared to White women (3% vs 1%, respectively; p < 0.0001). Residing in the South, being insured with Medicaid, and residing in a county with low median income were also associated with non-receipt of surgery. Black women remained less likely to undergo hysterectomy on multivariable analysis (OR 0.44, 95% CI 0.32-0.60). Non-receipt of hysterectomy was predictive of decreased CSS (HR 0.14, 95% CI 0.09-0.21) and OS (HR 0.18, 95% 0.14-0.23) on adjusted analysis. Black race was also an independent predictor of increased cancer-specific death (HR 2.07, 95% CI 1.50-2.86) as well as death from any cause (HR 1.74, 95% CI 1.44-2.09) on adjusted analysis. CONCLUSIONS: Black women with low-risk endometrial cancer were less likely to undergo hysterectomy and experienced decreased survival relative to White women. Further investigation is warranted to better understand the socioeconomic, geographic, and biologic factors that influence this disparity.

Molecular characterisation of a cellular conveyor belt in Clytia medusae.

The tentacular system of Clytia hemisphaerica medusa (Cnidaria, Hydrozoa) has recently emerged as a promising experimental model to tackle the developmental mechanisms that regulate cell lineage progression in an early-diverging animal phylum. From a population of proximal stem cells, the successive steps of tentacle stinging cell (nematocyte) elaboration, are spatially ordered along a "cellular conveyor belt". Furthermore, the C. hemisphaerica tentacular system exhibits bilateral organisation, with two perpendicular polarity axes (proximo-distal and oral-aboral). We aimed to improve our knowledge of this cellular system by combining RNAseq-based differential gene expression analyses and expression studies of Wnt signalling genes. RNAseq comparisons of gene expression levels were performed (i) between the tentacular system and a control medusa deprived of all tentacles, nematogenic sites and gonads, and (ii) between three samples staggered along the cellular conveyor belt. The behaviour in these differential expression analyses of two reference gene sets (stem cell genes; nematocyte genes), as well as the relative representations of selected gene ontology categories, support the validity of the cellular conveyor belt model. Expression patterns obtained by in situ hybridisation for selected highly differentially expressed genes and for Wnt signalling genes are largely consistent with the results from RNAseq. Wnt signalling genes exhibit complex spatial deployment along both polarity axes of the tentacular system, with the Wnt/ß-catenin pathway probably acting along the oral-aboral axis rather than the proximo-distal axis. These findings reinforce the idea that, despite overall radial symmetry, cnidarians have a full potential for elaboration of bilateral structures based on finely orchestrated deployment of an ancient developmental gene toolkit.

A software tool 'CroCo' detects pervasive cross-species contamination in next generation sequencing data.

BACKGROUND: Multiple RNA samples are frequently processed together and often mixed before multiplex sequencing in the same sequencing run. While different samples can be separated post sequencing using sample barcodes, the possibility of cross contamination between biological samples from different species that have been processed or sequenced in parallel has the potential to be extremely deleterious for downstream analyses. RESULTS: We present CroCo, a software package for identifying and removing such cross contaminants from assembled transcriptomes. Using multiple, recently published sequence datasets, we show that cross contamination is consistently present at varying levels in real data. Using real and simulated data, we demonstrate that CroCo detects contaminants efficiently and correctly. Using a real example from a molecular phylogenetic dataset, we show that contaminants, if not eliminated, can have a decisive, deleterious impact on downstream comparative analyses. CONCLUSIONS: Cross contamination is pervasive in new and published datasets and, if undetected, can have serious deleterious effects on downstream analyses. CroCo is a database-independent, multi-platform tool, designed for ease of use, that efficiently and accurately detects and removes cross contamination in assembled transcriptomes to avoid these problems. We suggest that the use of CroCo should become a standard cleaning step when processing multiple samples for transcriptome sequencing.

The ancestral gene repertoire of animal stem cells.

Stem cells are pivotal for development and tissue homeostasis of multicellular animals, and the quest for a gene toolkit associated with the emergence of stem cells in a common ancestor of all metazoans remains a major challenge for evolutionary biology. We reconstructed the conserved gene repertoire of animal stem cells by transcriptomic profiling of totipotent archeocytes in the demosponge Ephydatia fluviatilis and by tracing shared molecular signatures with flatworm and Hydra stem cells. Phylostratigraphy analyses indicated that most of these stem-cell genes predate animal origin, with only few metazoan innovations, notably including several partners of the Piwi machinery known to promote genome stability. The ancestral stem-cell transcriptome is strikingly poor in transcription factors. Instead, it is rich in RNA regulatory actors, including components of the "germ-line multipotency program" and many RNA-binding proteins known as critical regulators of mammalian embryonic stem cells.

How do environmental factors influence life cycles and development? An experimental framework for early-diverging metazoans.

Ecological developmental biology (eco-devo) explores the mechanistic relationships between the processes of individual development and environmental factors. Recent studies imply that some of these relationships have deep evolutionary origins, and may even pre-date the divergences of the simplest extant animals, including cnidarians and sponges. Development of these early diverging metazoans is often sensitive to environmental factors, and these interactions occur in the context of conserved signaling pathways and mechanisms of tissue homeostasis whose detailed molecular logic remain elusive. Efficient methods for transgenesis in cnidarians together with the ease of experimental manipulation in cnidarians and sponges make them ideal models for understanding causal relationships between environmental factors and developmental mechanisms. Here, we identify major questions at the interface between animal evolution and development and outline a road map for research aimed at identifying the mechanisms that link environmental factors to developmental mechanisms in early diverging metazoans. Also watch the Video Abstract.

Morphology and distribution of sensilla on head appendages in the water beetle Hygrobia hermanni (Coleoptera: Adephaga: Hygrobiidae).

Sensilla on head appendages were studied in detail for the first time in a member of the relict family Hygrobiidae (squeak beetles), closely related to Dytiscidae (diving beetles). Adult and third instar larval stage specimens of Hygrobia hermanni (Fabricius, 1775) were examined using scanning electron microscopy, focusing on antennae, palps and larval mandibles. In total, 37 sensilla subtypes are described, including 22 observed in the adult (basiconica: 3; Böhm's bristles: 2; circumvallate sensilla: 2; coeloconica: 10; ovoid placodea: 3; digitiform placodea: 2) and 16 in the larva (basiconica: 4; campaniformia: 1; chaetica: 4; coeloconica: 5; trichodea: 1; unnamed: 1). Only one subtype (of sensilla coeloconica) was shared between the adult and the larva. Autapomorphies of Hygrobiidae and Dytiscidae, and putatively shared derived characters (synapomorphies) of Hygrobiidae + Dytiscidae are discussed. Among the latter, the most remarkable is the acquisition of a special sensory field, located on the apical segment of the adult maxillary palp, subapically and postero-dorsally. This sensory field is made up of ovoid multiporous sensilla placodea otherwise present on the anterior (internal) surface of antennal segments, suggesting that in a common ancestor of Hygrobiidae and Dytiscidae, maxillary palps might have taken over enhanced capacities of longe-range molecule detection.

Maternally localized germ plasm mRNAs and germ cell/stem cell formation in the cnidarian Clytia.

The separation of the germ line from the soma is a classic concept in animal biology, and depending on species is thought to involve fate determination either by maternally localized germ plasm ("preformation" or "maternal inheritance") or by inductive signaling (classically termed "epigenesis" or "zygotic induction"). The latter mechanism is generally considered to operate in non-bilaterian organisms such as cnidarians and sponges, in which germ cell fate is determined at adult stages from multipotent stem cells. We have found in the hydrozoan cnidarian Clytia hemisphaerica that the multipotent "interstitial" cells (i-cells) in larvae and adult medusae, from which germ cells derive, express a set of conserved germ cell markers: Vasa, Nanos1, Piwi and PL10. In situ hybridization analyses unexpectedly revealed maternal mRNAs for all these genes highly concentrated in a germ plasm-like region at the egg animal pole and inherited by the i-cell lineage, strongly suggesting i-cell fate determination by inheritance of animal-localized factors. On the other hand, experimental tests showed that i-cells can form by epigenetic mechanisms in Clytia, since larvae derived from both animal and vegetal blastomeres separated during cleavage stages developed equivalent i-cell populations. Thus Clytia embryos appear to have maternal germ plasm inherited by i-cells but also the potential to form these cells by zygotic induction. Reassessment of available data indicates that maternally localized germ plasm molecular components were plausibly present in the common cnidarian/bilaterian ancestor, but that their role may not have been strictly deterministic.

Clytia hemisphaerica: a jellyfish cousin joins the laboratory.

Clytia hemisphaerica, a member of the early-branching animal phylum Cnidaria, is emerging rapidly as an experimental model for studies in developmental biology and evolution. Unlike the two existing genome-sequenced cnidarian models Nematostella and Hydra, Clytia has a free-swimming jellyfish form, which like "higher" animals (the Bilateria) has a complex organization including striated musculature, specialized nervous system and structured sensory and reproductive organs. Clytia has proved well suited to laboratory culture and to gene function analysis during early development. Initial studies have shed light on the origins of embryonic polarity and of the nematocyte as a specialized neurosensory cell, and on the regulation of oocyte maturation. With a full genome sequence soon to become available, and a clear potential for genetic approaches, Clytia is well placed to provide invaluable information on core mechanisms in cell and developmental biology, and on the evolution of key features of animal body plans.

New tricks with old genes: the genetic bases of novel cnidarian traits.

Recent thought on genome evolution has focused on the creation of new genes and changes in regulatory mechanisms while ignoring the role of selective gene loss in shaping genomes. Using data from two cnidarians, the jellyfish Clytia and the coral Acropora, we examined the relative significance of new 'taxonomically restricted' genes and selectively retained ancestral genes in enabling the evolution of novel traits. Consistent with its more complex life-cycle, the proportion of novel genes identified in Clytia was higher than that in the 'polyp only' cnidarians Nematostella and Hydra, but each of these cnidarians has retained a proportion of ancestral genes not present in the other two. The ubiquity and near-stochastic nature of gene loss can explain the discord between patterns of gene distribution and taxonomy.

A new semi-subterranean diving beetle of the Hydroporus normandi-complex from south-eastern France, with notes on other taxa of the complex (Coleoptera: Dytiscidae).

Hydroporus galloprovincialis sp. n. is described from Jouques, north-east of Aix-en-Provence in south-eastern France (Provence). The new species belongs to the Hydroporus memnonius-group, within which it is the only ascertained normandi-complex species of the French fauna. The habitat (small springs) and morphological features of the species suggest a semi-subterranean life style. It can be separated from the other species of the complex notably by its more elongated and parallel-sided habitus, by its wider pronotum and flatter dorsal surface, and furthermore differs from H. normandi Régimbart, 1903 by a different shape of the gonocoxa. Illustrations are provided for other described normandi-complex species and subspecies for comparison. Analyses of partial cox1 sequences indicate that the three sampled H. normandi-complex species (H. galloprovincialis sp. n., H. normandi and H. lluci Fery, 1999) diverged recently; their relationships are unresolved. H. galloprovincialis sp. n. is for the moment the only species of the genus Hydroporus Clairville, 1806 endemic to continental France.

Methodological evaluation of original articles on radiomics and machine learning for outcome prediction based on positron emission tomography (PET).

AIM: Despite a vast number of articles on radiomics and machine learning in positron emission tomography (PET) imaging, clinical applicability remains limited, partly owing to poor methodological quality. We therefore systematically investigated the methodology described in publications on radiomics and machine learning for PET-based outcome prediction. METHODS: A systematic search for original articles was run on PubMed. All articles were rated according to 17 criteria proposed by the authors. Criteria with >2 rating categories were binarized into "adequate" or "inadequate". The association between the number of "adequate" criteria per article and the date of publication was examined. RESULTS: One hundred articles were identified (published between 07/2017 and 09/2023). The median proportion of articles per criterion that were rated "adequate" was 65% (range: 23-98%). Nineteen articles (19%) mentioned neither a test cohort nor cross-validation to separate training from testing. The median number of criteria with an "adequate" rating per article was 12.5 out of 17 (range, 4-17), and this did not increase with later dates of publication (Spearman's rho, 0.094; p = 0.35). In 22 articles (22%), less than half of the items were rated "adequate". Only 8% of articles published the source code, and 10% made the dataset openly available. CONCLUSION: Among the articles investigated, methodological weaknesses have been identified, and the degree of compliance with recommendations on methodological quality and reporting shows potential for improvement. Better adherence to established guidelines could increase the clinical significance of radiomics and machine learning for PET-based outcome prediction and finally lead to the widespread use in routine clinical practice.

Artificial Intelligence-powered automatic volume calculation in medical images - available tools, performance and challenges for nuclear medicine.

Volumetry is crucial in oncology and endocrinology, for diagnosis, treatment planning, and evaluating response to therapy for several diseases. The integration of Artificial Intelligence (AI) and Deep Learning (DL) has significantly accelerated the automatization of volumetric calculations, enhancing accuracy and reducing variability and labor. In this review, we show that a high correlation has been observed between Machine Learning (ML) methods and expert assessments in tumor volumetry; Yet, it is recognized as more challenging than organ volumetry. Liver volumetry has shown progression in accuracy with a decrease in error. If a relative error below 10 % is acceptable, ML-based liver volumetry can be considered reliable for standardized imaging protocols if used in patients without major anomalies. Similarly, ML-supported automatic kidney volumetry has also shown consistency and reliability in volumetric calculations. In contrast, AI-supported thyroid volumetry has not been extensively developed, despite initial works in 3D ultrasound showing promising results in terms of accuracy and reproducibility. Despite the advancements presented in the reviewed literature, the lack of standardization limits the generalizability of ML methods across diverse scenarios. The domain gap, i. e., the difference in probability distribution of training and inference data, is of paramount importance before clinical deployment of AI, to maintain accuracy and reliability in patient care. The increasing availability of improved segmentation tools is expected to further incorporate AI methods into routine workflows where volumetry will play a more prominent role in radionuclide therapy planning and quantitative follow-up of disease evolution.

The Medical Informatics Initiative and the Network University Medicine - Perspectives for Nuclear Medicine.

Digitization in the healthcare sector and the support of clinical workflows with artificial intelligence (AI), including AI-supported image analysis, represent a great challenge and equally a promising perspective for preclinical and clinical nuclear medicine. In Germany, the Medical Informatics Initiative (MII) and the Network University Medicine (NUM) are of central importance for this transformation. This review article outlines these structures and highlights their future role in enabling privacy-preserving federated multi-center analyses with interoperable data structures harmonized between site-specific IT infrastructures. The newly founded working group "Digitization and AI" in the German Society of Nuclear Medicine (DGN) as well as the Fach- und Organspezifische Arbeitsgruppe (FOSA, specialty- and organ-specific working group) founded for the field of nuclear medicine (FOSA Nuklearmedizin) within the NUM aim to initiate and coordinate measures in the context of digital medicine and (image-)data-driven analyses for the DGN.

Somatic stem cells express Piwi and Vasa genes in an adult ctenophore: ancient association of "germline genes" with stemness.

Stem cells are essential for animal development and adult tissue homeostasis, and the quest for an ancestral gene fingerprint of stemness is a major challenge for evolutionary developmental biology. Recent studies have indicated that a series of genes, including the transposon silencer Piwi and the translational activator Vasa, specifically involved in germline determination and maintenance in classical bilaterian models (e.g., vertebrates, fly, nematode), are more generally expressed in adult multipotent stem cells in other animals like flatworms and hydras. Since the progeny of these multipotent stem cells includes both somatic and germinal derivatives, it remains unclear whether Vasa, Piwi, and associated genes like Bruno and PL10 were ancestrally linked to stemness, or to germinal potential. We have investigated the expression of Vasa, two Piwi paralogues, Bruno and PL10 in Pleurobrachia pileus, a member of the early-diverging phylum Ctenophora, the probable sister group of cnidarians. These genes were all expressed in the male and female germlines, and with the exception of one of the Piwi paralogues, they showed similar expression patterns within somatic territories (tentacle root, comb rows, aboral sensory complex). Cytological observations and EdU DNA-labelling and long-term retention experiments revealed concentrations of stem cells closely matching these gene expression areas. These stem cell pools are spatially restricted, and each specialised in the production of particular types of somatic cells. These data unveil important aspects of cell renewal within the ctenophore body and suggest that Piwi, Vasa, Bruno, and PL10 belong to a gene network ancestrally acting in two distinct contexts: (i) the germline and (ii) stem cells, whatever the nature of their progeny.

Independent specialisation of myosin II paralogues in muscle vs. non-muscle functions during early animal evolution: a ctenophore perspective.

BACKGROUND: Myosin II (or Myosin Heavy Chain II, MHCII) is a family of molecular motors involved in the contractile activity of animal muscle cells but also in various other cellular processes in non-muscle cells. Previous phylogenetic analyses of bilaterian MHCII genes identified two main clades associated respectively with smooth/non-muscle cells (MHCIIa) and striated muscle cells (MHCIIb). Muscle cells are generally thought to have originated only once in ancient animal history, and decisive insights about their early evolution are expected to come from expression studies of Myosin II genes in the two non-bilaterian phyla that possess muscles, the Cnidaria and Ctenophora. RESULTS: We have uncovered three MHCII paralogues in the ctenophore species Pleurobrachia pileus. Phylogenetic analyses indicate that the MHCIIa / MHCIIb duplication is more ancient than the divergence between extant metazoan lineages. The ctenophore MHCIIa gene (PpiMHCIIa) has an expression pattern akin to that of "stem cell markers" (Piwi, Vasa…) and is expressed in proliferating cells. We identified two MHCIIb genes that originated from a ctenophore-specific duplication. PpiMHCIIb1 represents the exclusively muscular form of myosin II in ctenophore, while PpiMHCIIb2 is expressed in non-muscle cells of various types. In parallel, our phalloidin staining and TEM observations highlight the structural complexity of ctenophore musculature and emphasize the experimental interest of the ctenophore tentacle root, in which myogenesis is spatially ordered and strikingly similar to striated muscle formation in vertebrates. CONCLUSION: MHCIIa expression in putative stem cells/proliferating cells probably represents an ancestral trait, while specific involvement of some MHCIIa genes in smooth muscle fibres is a uniquely derived feature of the vertebrates. That one ctenophore MHCIIb paralogue (PpiMHCIIb2) has retained MHCIIa-like expression features furthermore suggests that muscular expression of the other paralogue, PpiMHCIIb1, was the result of neofunctionalisation within the ctenophore lineage, making independent origin of ctenophore muscle cells a likely option.

Convergent origins and rapid evolution of spliced leader trans-splicing in metazoa: insights from the ctenophora and hydrozoa.

Replacement of mRNA 5' UTR sequences by short sequences trans-spliced from specialized, noncoding, spliced leader (SL) RNAs is an enigmatic phenomenon, occurring in a set of distantly related animal groups including urochordates, nematodes, flatworms, and hydra, as well as in Euglenozoa and dinoflagellates. Whether SL trans-splicing has a common evolutionary origin and biological function among different organisms remains unclear. We have undertaken a systematic identification of SL exons in cDNA sequence data sets from non-bilaterian metazoan species and their closest unicellular relatives. SL exons were identified in ctenophores and in hydrozoan cnidarians, but not in other cnidarians, placozoans, or sponges, or in animal unicellular relatives. Mapping of SL absence/presence obtained from this and previous studies onto current phylogenetic trees favors an evolutionary scenario involving multiple origins for SLs during eumetazoan evolution rather than loss from a common ancestor. In both ctenophore and hydrozoan species, multiple SL sequences were identified, showing high sequence diversity. Detailed analysis of a large data set generated for the hydrozoan Clytia hemisphaerica revealed trans-splicing of given mRNAs by multiple alternative SLs. No evidence was found for a common identity of trans-spliced mRNAs between different hydrozoans. One feature found specifically to characterize SL-spliced mRNAs in hydrozoans, however, was a marked adenosine enrichment immediately 3' of the SL acceptor splice site. Our findings of high sequence divergence and apparently indiscriminate use of SLs in hydrozoans, along with recent findings in other taxa, indicate that SL genes have evolved rapidly in parallel in diverse animal groups, with constraint on SL exon sequence evolution being apparently rare.

Homology of arthropod anterior appendages revealed by Hox gene expression in a sea spider.

Arthropod head segments offer a paradigm for understanding the diversification of form during evolution, as a variety of morphologically diverse appendages have arisen from them. There has been long-running controversy, however, concerning which head appendages are homologous among arthropods, and from which ancestral arrangement they have been derived. This controversy has recently been rekindled by the proposition that the probable ancestral arrangement, with appendages on the first head segment, has not been lost in all extant arthropods as previously thought, but has been retained in the pycnogonids, or sea spiders. This proposal was based on the neuroanatomical analysis of larvae from the sea spider Anoplodactylus sp., and suggested that the most anterior pair of appendages, the chelifores, are innervated from the first part of the brain, the protocerebrum. Our examination of Hox gene expression in another sea spider, Endeis spinosa, refutes this hypothesis. The anterior boundaries of Hox gene expression domains place the chelifore appendages as clearly belonging to the second head segment, innervated from the second part of the brain, the deutocerebrum. The deutocerebrum must have been secondarily displaced towards the protocerebrum in pycnogonid ancestors. As anterior-most appendages are also deutocerebral in the other two arthropod groups, the Euchelicerata and the Mandibulata, we conclude that the protocerebral appendages have been lost in all extant arthropods.

The aquatic Adephaga of the Makay, central-western Madagascar, with description of two new diving beetle species (Coleoptera, Gyrinidae, Haliplidae, Noteridae, Dytiscidae).

Water beetles of the families Gyrinidae, Haliplidae, Noteridae, and Dytiscidae (aquatic Adephaga) of the Makay in central-western Madagascar were surveyed in three campaigns during the years 2016-2018. A total of 74 species was collected from 62 sampling sites, all except one being newly recorded from the Makay. Copelatusmalavergnorum sp. nov. (irinus group) and C.zanabato sp. nov. (erichsonii group) (Dytiscidae, Copelatinae) are described and their habitus and male genitalia are illustrated. A systematic account is given, including description of habitat preferences for each species. Analyses of species composition and dominance, species diversity and endemism highlighted the strong singularity of the aquatic Adephaga fauna inhabiting the sandstone massif of inner Makay (notably with several local endemic dytiscids) with respect to its peripheral lowlands. These comparisons were also performed between groups of sites categorised according to vegetation context (forested, semi-forested, non-forested). Rather unexpectedly, inner Makay although well-preserved and little deforested has relatively low endemism level and low species diversity (H1 Hill number twice lower than in the geographically close and geologically similar massif of Isalo). Species diversity was higher in the deforested and man-impacted peripheral sites, which yielded a rich contingent of western Madagascar lowland species including a few undescribed or rarely observed dytiscids.

Inferring FDG-PET-positivity of lymph node metastases in proven lung cancer from contrast-enhanced CT using radiomics and machine learning.

BACKGROUND: We evaluated the role of radiomics applied to contrast-enhanced computed tomography (CT) in the detection of lymph node (LN) metastases in patients with known lung cancer compared to 18F-fluorodeoxyglucose positron emission tomography (PET)/CT as a reference. METHODS: This retrospective analysis included 381 patients with 1,799 lymph nodes (450 malignant, 1,349 negative). The data set was divided into a training and validation set. A radiomics analysis with 4 filters and 6 algorithms resulting in 24 different radiomics signatures and a bootstrap algorithm (Bagging) with 30 bootstrap iterations was performed. A decision curve analysis was applied to generate a net benefit to compare the radiomics signature to two expert radiologists as one-by-one and as a prescreening tool in combination with the respective radiologist and only the radiologists. RESULTS: All 24 modeling methods showed good and reliable discrimination for malignant/benign LNs (area under the curve 0.75-0.87). The decision curve analysis showed a net benefit for the least absolute shrinkage and selection operator (LASSO) classifier for the entire probability range and outperformed the expert radiologists except for the high probability range. Using the radiomics signature as a prescreening tool for the radiologists did not improve net benefit. CONCLUSIONS: Radiomics showed good discrimination power irrespective of the modeling technique in detecting LN metastases in patients with known lung cancer. The LASSO classifier was a suitable diagnostic tool and even outperformed the expert radiologists, except for high probabilities. Radiomics failed to improve clinical benefit as a prescreening tool.