RESUMO
OBJECTIVE: The aim of this study was to develop and validate an interpretable and highly generalizable multimodal radiomics model for predicting the prognosis of patients with cerebral hemorrhage. METHODS: This retrospective study involved 237 patients with cerebral hemorrhage from 3 medical centers, of which a training cohort of 186 patients (medical center 1) was selected and 51 patients from medical center 2 and medical center 3 were used as an external testing cohort. A total of 1762 radiomics features were extracted from nonenhanced computed tomography using Pyradiomics, and the relevant macroscopic imaging features and clinical factors were evaluated by 2 experienced radiologists. A radiomics model was established based on radiomics features using the random forest algorithm, and a radiomics-clinical model was further trained by combining radiomics features, clinical factors, and macroscopic imaging features. The performance of the models was evaluated using area under the curve (AUC), sensitivity, specificity, and calibration curves. Additionally, a novel SHAP (SHAPley Additive exPlanations) method was used to provide quantitative interpretability analysis for the optimal model. RESULTS: The radiomics-clinical model demonstrated superior predictive performance overall, with an AUC of 0.88 (95% confidence interval, 0.76-0.95; P < 0.01). Compared with the radiomics model (AUC, 0.85; 95% confidence interval, 0.72-0.94; P < 0.01), there was a 0.03 improvement in AUC. Furthermore, SHAP analysis revealed that the fusion features, rad score and clinical rad score, made significant contributions to the model's decision-making process. CONCLUSION: Both proposed prognostic models for cerebral hemorrhage demonstrated high predictive levels, and the addition of macroscopic imaging features effectively improved the prognostic ability of the radiomics-clinical model. The radiomics-clinical model provides a higher level of predictive performance and model decision-making basis for the risk prognosis of cerebral hemorrhage.
RESUMO
OBJECTIVES: The purpose of this study is to inquire about the potential association between radiomics features and the pathological nature of thyroid nodules (TNs), and to propose an interpretable radiomics-based model for predicting the risk of malignant TN. METHODS: In this retrospective study, computed tomography (CT) imaging and pathological data from 141 patients with TN were collected. The data were randomly stratified into a training group (n = 112) and a validation group (n = 29) at a ratio of 4:1. A total of 1316 radiomics features were extracted by using the pyradiomics tool. The redundant features were removed through correlation testing, and the least absolute shrinkage and selection operator (LASSO) or the minimum redundancy maximum relevance standard was used to select features. Finally, 4 different machine learning models (RF Hybrid Feature, SVM Hybrid Feature, RF, and LASSO) were constructed. The performance of the 4 models was evaluated using the receiver operating characteristic curve. The calibration curve, decision curve analysis, and SHapley Additive exPlanations method were used to evaluate or explain the best radiomics machine learning model. RESULTS: The optimal radiomics model (RF Hybrid Feature model) demonstrated a relatively high degree of discrimination with an area under the receiver operating characteristic curve (AUC) of 0.87 (95% CI, 0.70-0.97; P < 0.001) for the validation cohort. Compared with the commonly used LASSO model (AUC, 0.78; 95% CI, 0.60-0.91; P < 0.01), there is a significant improvement in AUC in the validation set, net reclassification improvement, 0.79 (95% CI, 0.13-1.46; P < 0.05), and integrated discrimination improvement, 0. 20 (95% CI, 0.10-0.30; P < 0.001). CONCLUSION: The interpretable radiomics model based on CT performs well in predicting benign and malignant TNs by using quantitative radiomics features of the unilateral total thyroid. In addition, the data preprocessing method incorporating different layers of features has achieved excellent experimental results. CLINICAL RELEVANCE STATEMENT: As the detection rate of TNs continues to increase, so does the diagnostic burden on radiologists. This study establishes a noninvasive, interpretable and accurate machine learning model to rapidly identify the nature of TN found in CT.
Assuntos
Bócio Nodular , Nódulo da Glândula Tireoide , Humanos , Radiômica , Estudos Retrospectivos , Nódulo da Glândula Tireoide/diagnóstico por imagemRESUMO
OBJECTIVES: To develop and validate a novel interpretable artificial intelligence (AI) model that integrates radiomic features, deep learning features, and imaging features at multiple semantic levels to predict the prognosis of intracerebral hemorrhage (ICH) patients at 6 months post-onset. MATERIALS AND METHODS: Retrospectively enrolled 222 patients with ICH for Non-contrast Computed Tomography (NCCT) images and clinical data, who were divided into a training cohort (n = 186, medical center 1) and an external testing cohort (n = 36, medical center 2). Following image preprocessing, the entire hematoma region was segmented by two radiologists as the volume of interest (VOI). Pyradiomics algorithm library was utilized to extract 1762 radiomics features, while a deep convolutional neural network (EfficientnetV2-L) was employed to extract 1000 deep learning features. Additionally, radiologists evaluated imaging features. Based on the three different modalities of features mentioned above, the Random Forest (RF) model was trained, resulting in three models (Radiomics Model, Radiomics-Clinical Model, and DL-Radiomics-Clinical Model). The performance and clinical utility of the models were assessed using the Area Under the Receiver Operating Characteristic Curve (AUC), calibration curve, and Decision Curve Analysis (DCA), with AUC compared using the DeLong test. Furthermore, this study employs three methods, Shapley Additive Explanations (SHAP), Grad-CAM, and Guided Grad-CAM, to conduct a multidimensional interpretability analysis of model decisions. RESULTS: The Radiomics-Clinical Model and DL-Radiomics-Clinical Model exhibited relatively good predictive performance, with an AUC of 0.86 [95% Confidence Intervals (CI): 0.71, 0.95; P < 0.01] and 0.89 (95% CI: 0.74, 0.97; P < 0.01), respectively, in the external testing cohort. CONCLUSION: The multimodal explainable AI model proposed in this study can accurately predict the prognosis of ICH. Interpretability methods such as SHAP, Grad-CAM, and Guided Grad-Cam partially address the interpretability limitations of AI models. Integrating multimodal imaging features can effectively improve the performance of the model. CLINICAL RELEVANCE STATEMENT: Predicting the prognosis of patients with ICH is a key objective in emergency care. Accurate and efficient prognostic tools can effectively prevent, manage, and monitor adverse events in ICH patients, maximizing treatment outcomes.
Assuntos
Inteligência Artificial , Hemorragia Cerebral , Aprendizado Profundo , Tomografia Computadorizada por Raios X , Humanos , Hemorragia Cerebral/diagnóstico por imagem , Prognóstico , Tomografia Computadorizada por Raios X/métodos , Masculino , Feminino , Estudos Retrospectivos , Pessoa de Meia-Idade , Idoso , Curva ROC , Redes Neurais de Computação , AlgoritmosRESUMO
Copper species were incorporated into SBA-15 by solid-state grinding precursor with as-prepared mesoporous silica (SPA). The obtained materials (CuAS) were well-characterized by XRD, TEM, N(2) adsorption, H(2)-TPR, IR, and TG and compared with the material derived from calcined SBA-15 (CuCS). Surprisingly, CuO up to 6.7 mmol·g(-1) can be highly dispersed on SBA-15 by use of SPA strategy. Such CuO forms a smooth layer coated on the internal walls of SBA-15, which contributes to the spatial order and results in less-blocked mesopores. However, the aggregation of CuO takes place in CuCS material containing 6.7 mmol·g(-1) copper, which generates large CuO particles of 21.4 nm outside the mesopores. We reveal that the high dispersion extent of CuO is ascribed to the abundant silanols, as well as the confined space between template and silica walls provided by as-prepared SBA-15. The SPA strategy allows template removal and precursor conversion in one step, avoids the repeated calcination in conventional modification process, and saves time and energy. We also demonstrate that the CuAS material after autoreduction exhibits much better adsorptive desulfurization capacity than CuCS. Moreover, the adsorption capacity of regenerated adsorbent can be recovered completely.
RESUMO
The middle and lower portions of the Yangtze River basin is the most species-rich region for freshwater mussels in Asia. The management and conservation of the taxa in this region has been greatly hampered by the lack of a well-developed phylogeny and species-level taxonomic framework. In this study, we tested the utility of two mitochondrial genes commonly used as DNA barcodes: the first subunit of the cytochrome oxidase c gene (COI) and the first subunit of the NADH dehydrogenase gene (ND1) for 34 putative species representing 15 genera, and also generated phylogenetic hypotheses for Chinese unionids based on the combined dataset of the two mitochondrial genes. The results showed that both loci performed well as barcodes for species identification, but the ND1 sequences provided better resolution when compared to COI. Based on the two-locus dataset, Bayesian Inference (BI) and Maximum Likelihood (ML) phylogenetic analyses indicated 3 of the 15 genera of Chinese freshwater mussels examined were polyphyletic. Additionally, the analyses placed the 15 genera into 3 subfamilies: Unioninae (Aculamprotula, Cuneopsis, Nodularia and Schistodesmus), Gonideninae (Lamprotula, Solenaia and Ptychorhychus) and Anodontinae (Cristaria, Arconaia, Acuticosta, Lanceolaria, Anemina and Sinoanodonta). Our results contradict previous taxonomic classification that placed the genera Arconaia, Acuticosta and Lanceolaria in the Unioninae. This study represents one of the first attempts to develop a molecular phylogenetic framework for the Chinese members of the Unionidae and will provide a basis for future research on the evolution, ecology, and conservation of Chinese freshwater mussels.
Assuntos
Bivalves/genética , Código de Barras de DNA Taxonômico/métodos , Animais , China , DNA Mitocondrial/genética , Complexo IV da Cadeia de Transporte de Elétrons/genética , Água Doce , Genes Mitocondriais , Especiação Genética , Genoma Mitocondrial , NADH Desidrogenase/genética , Filogenia , Rios , Unionidae/genéticaRESUMO
Acorssocheilus beijiangensis is an endemic south China stream-dwelling cyprinid species. Its complete mitochondrial genome is 16,596 bp in length, consisting of 13 protein-coding genes, 22 tRNA genes (ranging from 67 bp in tRNACys to 76 bp in tRNALeu and tRNALys ), two rRNA genes (959 bp in 12S rRNA and 1683 bp in 16S rRNA), and one control region (937 bp). Its overall base composition is A: 31.1%, C: 27.9%, G: 16.2%, and T: 124.8%. The complete mitogenome of the Chinese barred species of Cyprinidae could provide a basic data for further phylogenetics analysis.
RESUMO
Acrossocheilus jishouensis is an endemic south China stream-dwelling cyprinid species. Its complete mitochondrial genome is 16,587 bp in length, consisting of 13 protein-coding genes, 22 tRNA genes (ranging from 67 bp in tRNACys to 76 bp in tRNALeu and tRNALys ), two rRNA genes (956 bp in 12S rRNA and 1673 bp in 16S rRNA), and one control region (942 bp). Its overall base composition is A: 31.2%, C: 27.6%, G: 16.2%, and T: 25.1%. The complete mitogenome of the Chinese barred species of Cpynidae could provide a basic data for further phylogenetics analysis.
RESUMO
The taxonomy of genus Anodonta is rather ambiguous, as it has great variation on the shell shape. Anodonta lucida is an endemic species of freshwater mussel in China, characterized by shining epidermis. The complete maternal mitochondrial genome of freshwater mussel A. lucida was first determined (GenBank accession no. KF667529). The genome is 16,285 bp long with an AT content of 64.02%. All the 37 typical animal mitochondrial genes are found, including 13 protein-coding genes, 22 tRNA genes, and 2 rRNA genes. The genome also contains 24 unassigned regions, ranking from 1 to 830 bp in length, the largest of which is the putative control region (CR). The base composition of the genome is A (36.32%), G (13.01%), T (27.70%) and C (22.98%). Gene order is identical to other species of Unionidae except Gonideinae.
Assuntos
Genoma Mitocondrial/genética , Análise de Sequência de DNA , Unionidae/genética , Animais , Água Doce , Genes de RNAr/genética , Anotação de Sequência Molecular , Dados de Sequência Molecular , Fases de Leitura Aberta/genética , RNA de Transferência/genéticaRESUMO
Isolated lithium sites were anchored on mesoporous silica by a molecular precursor approach at room temperature. The resultant materials exhibit ordered mesostructure, high base strength, and more importantly, a molecular-level dispersion of active sites, which are extremely desirable for catalysis and impossible to be realized by conventional methods.
RESUMO
A novel π-complexation adsorbent is fabricated by grafting Cu(I)-containing molecule precursors onto ß-cyclodextrin. The adsorbent provides a molecular-level dispersion of Cu(I), which is particularly beneficial to the adsorptive removal of aromatic sulfur thiophene, and is impossible to be realized through the conventional thermal method.