Computational and structural biotechnology journal最新文献

{"title":"Advancing titanium dioxide coated photocatalytic depolluting surfaces: Leveraging ASINA's roadmap for safer and sustainable solutions","authors":"Irini Furxhi ,&nbsp;Massimo Perucca ,&nbsp;Giovanni Baldi ,&nbsp;Valentina Dami ,&nbsp;Andrea Cioni ,&nbsp;Antti Joonas Koivisto ,&nbsp;Rossella Bengalli ,&nbsp;Paride Mantecca ,&nbsp;Giulia Motta ,&nbsp;Marie Carriere ,&nbsp;Alessia Nicosia ,&nbsp;Fabrizio Ravegnani ,&nbsp;David Burrueco-Subirà ,&nbsp;Socorro Vázquez-Campos ,&nbsp;Elma Lahive ,&nbsp;Jesús Lopez de Ipiña ,&nbsp;Juliana Oliveira ,&nbsp;Patrick Cronin ,&nbsp;Magda Blosi ,&nbsp;Anna Costa","doi":"10.1016/j.csbj.2024.10.001","DOIUrl":"10.1016/j.csbj.2024.10.001","url":null,"abstract":"<div><div>This report, the second of its kind from ASINA project, aims at providing a roadmap with quantitative metrics for Safe(r) and (more) Sustainable by Design (SSbD) solutions for titanium dioxide (TiO₂) nanomaterials (NMs). We begin with a brief description of ASINA’s methodology across the product lifecycle, highlighting the quantitative elements, such as the Key Performance Indicators (KPIs). We then propose a decision support tool for implementing SSbD objectives across various dimensions—functionality, cost, environment, and human health safety. This is followed by the main innovative findings, a consolidation of the technical processes involved, design rationales, experimental procedures, tools and models, used and developed, to deliver photocatalytic depolluting surfaces by spray- finishing techniques based on TiO₂ NMs formulations. The roadmap is thoroughly described to inform similar projects through the integration of KPIs into SSbD methodologies, fostering data-driven decision-making. While specific results are beyond this report's scope, its primary aim is to demonstrate the roadmap (SSbD know-how) and promote SSbD-oriented innovation in nanotechnology. Finally, we provide a comparison of the approaches followed in two case studies that target different industrial sectors. This case-specific SSbD assessments provide a concrete exemplification of the addressed methodology that contributes to the efforts towards attaining a common roadmap for implementing SSbD solutions aligned with the EU’s Green Deal objectives.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"25 ","pages":"Pages 269-280"},"PeriodicalIF":4.4,"publicationDate":"2024-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142657183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"NanoBioAccumulate: Modelling the uptake and bioaccumulation of nanomaterials in soil and aquatic invertebrates via the Enalos DIAGONAL Cloud Platform","authors":"Dimitris G. Mintis ,&nbsp;Nikolaos Cheimarios ,&nbsp;Andreas Tsoumanis ,&nbsp;Anastasios G. Papadiamantis ,&nbsp;Nico W. van den Brink ,&nbsp;Henk J. van Lingen ,&nbsp;Georgia Melagraki ,&nbsp;Iseult Lynch ,&nbsp;Antreas Afantitis","doi":"10.1016/j.csbj.2024.09.028","DOIUrl":"10.1016/j.csbj.2024.09.028","url":null,"abstract":"<div><div><em>NanoBioAccumulate</em> is a free-to-use web-based tool hosted on the Enalos DIAGONAL Cloud Platform (<span><span>https://www.enaloscloud.novamechanics.com/diagonal/pbpk/</span><svg><path></path></svg></span>) that provides users with the capability to model and predict the uptake and bioaccumulation of nanomaterials (NMs) by soil and aquatic invertebrates using two common first-order one-compartment biokinetic models. <em>NanoBioAccumulate</em> offers an approach for comprehensively analyzing the kinetics of different forms of NMs via a nonlinear fitting feature, integrating them with environmental fate models, and considering important physiological processes. <em>NanoBioAccumulate</em> overcomes the constraint of requiring prior knowledge of kinetic rate constants associated with the biokinetic models and eliminates the need for external statistical analysis software as it quantifies the kinetic rate constants and other constants through the application of nonlinear regression, using user-provided experimental data. Furthermore, <em>NanoBioAccumulate</em> incorporates statistical analysis measures like the adjusted R-squared and the bias-corrected Akaike information criterion, allowing for assessment of the goodness-of-fit of the two different biokinetic models, assisting in the identification of the best-performing model for a specific nanoform and its uptake kinetics by a specific invertebrate. The tool also includes model scenarios, retrieved from literature, which involve examining the exposure of soil and aquatic invertebrates to various types of NMs such as TiO₂, SiO₂, C₆₀, graphene, graphene oxide (GO), Au, Ag and its ionic control AgNO₃. These model scenarios aim to enhance understanding of the uptake and elimination rates exhibited by different NM-species. <em>NanoBioAccumulate</em> features advanced integration capabilities, enabled by an extensive Application Programming Interface (API). This functionality promotes efficient data exchange and interoperability with other software and web applications, significantly expanding its utility in research, regulatory risk assessment and environmental surveillance and monitoring contexts. The inclusion of a user-friendly Graphical User Interface (GUI) in <em>NanoBioAccumulate</em> greatly improves the overall user experience by simplifying complex tasks and eliminating the need for programming proficiency, thereby expanding the tool's applicability to a diverse range of users across various fields such as environmental research, monitoring, and regulation.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"25 ","pages":"Pages 243-255"},"PeriodicalIF":4.4,"publicationDate":"2024-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"BioKGrapher: Initial evaluation of automated knowledge graph construction from biomedical literature","authors":"Henning Schäfer ,&nbsp;Ahmad Idrissi-Yaghir ,&nbsp;Kamyar Arzideh ,&nbsp;Hendrik Damm ,&nbsp;Tabea M.G. Pakull ,&nbsp;Cynthia S. Schmidt ,&nbsp;Mikel Bahn ,&nbsp;Georg Lodde ,&nbsp;Elisabeth Livingstone ,&nbsp;Dirk Schadendorf ,&nbsp;Felix Nensa ,&nbsp;Peter A. Horn ,&nbsp;Christoph M. Friedrich","doi":"10.1016/j.csbj.2024.10.017","DOIUrl":"10.1016/j.csbj.2024.10.017","url":null,"abstract":"<div><div><strong>Background</strong> The growth of biomedical literature presents challenges in extracting and structuring knowledge. Knowledge Graphs (KGs) offer a solution by representing relationships between biomedical entities. However, manual construction of KGs is labor-intensive and time-consuming, highlighting the need for automated methods. This work introduces BioKGrapher, a tool for automatic KG construction using large-scale publication data, with a focus on biomedical concepts related to specific medical conditions. BioKGrapher allows researchers to construct KGs from PubMed IDs.</div><div><strong>Methods</strong> The BioKGrapher pipeline begins with Named Entity Recognition and Linking (NER+NEL) to extract and normalize biomedical concepts from PubMed, mapping them to the Unified Medical Language System (UMLS). Extracted concepts are weighted and re-ranked using Kullback-Leibler divergence and local frequency balancing. These concepts are then integrated into hierarchical KGs, with relationships formed using terminologies like SNOMED CT and NCIt. Downstream applications include multi-label document classification using Adapter-infused Transformer models.</div><div><strong>Results</strong> BioKGrapher effectively aligns generated concepts with clinical practice guidelines from the German Guideline Program in Oncology (GGPO), achieving <span><math><msub><mrow><mi>F</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-Scores of up to 0.6. In multi-label classification, Adapter-infused models using a BioKGrapher cancer-specific KG improved micro <span><math><msub><mrow><mi>F</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>-Scores by up to 0.89 percentage points over a non-specific KG and 2.16 points over base models across three BERT variants. The drug-disease extraction case study identified indications for Nivolumab and Rituximab.</div><div><strong>Conclusion</strong> BioKGrapher is a tool for automatic KG construction, aligning with the GGPO and enhancing downstream task performance. It offers a scalable solution for managing biomedical knowledge, with potential applications in literature recommendation, decision support, and drug repurposing.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"24 ","pages":"Pages 639-660"},"PeriodicalIF":4.4,"publicationDate":"2024-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526863","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Toxicogenomic assessment of in vitro macrophages exposed to profibrotic challenge reveals a sustained transcriptomic immune signature","authors":"Jack Morikka ,&nbsp;Antonio Federico ,&nbsp;Lena Möbus ,&nbsp;Simo Inkala ,&nbsp;Alisa Pavel ,&nbsp;Saara Sani ,&nbsp;Maaret Vaani ,&nbsp;Sanna Peltola ,&nbsp;Angela Serra ,&nbsp;Dario Greco","doi":"10.1016/j.csbj.2024.10.010","DOIUrl":"10.1016/j.csbj.2024.10.010","url":null,"abstract":"<div><div>Immune signalling is a crucial component in the progression of fibrosis. However, approaches for the safety assessment of potentially profibrotic substances, that provide information on mechanistic immune responses, are underdeveloped. This study aimed to develop a novel framework for assessing the immunotoxicity of fibrotic compounds. We exposed macrophages in vitro to multiple sublethal concentrations of the profibrotic agent bleomycin, over multiple timepoints, and generated RNA sequencing data. Using a toxicogenomic approach, we performed dose-dependent analysis to discover genes dysregulated by bleomycin exposure in a dose-responsive manner. A subset of immune genes displayed a sustained dose-dependent and differential expression response to profibrotic challenge. An immunoassay revealed cytokines and proteinases responding to bleomycin exposure that closely correlated to transcriptomic alterations, underscoring the integration between transcriptional immune response and external immune signalling activity. This study not only increases our understanding of the immunological mechanisms of fibrosis, but also offers an innovative framework for the toxicological evaluation of substances with potential fibrogenic effects on macrophage signalling. Our work brings a new immunotoxicogenomic direction for hazard assessment of fibrotic compounds, through the implementation of a time and resource efficient in vitro methodology.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"25 ","pages":"Pages 194-204"},"PeriodicalIF":4.4,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transforming in-clinic post-operative and intermediate care with cosinuss°","authors":"Incinur Zellhuber ,&nbsp;Melanie Schade ,&nbsp;Tim Adams ,&nbsp;Manfred Blobner ,&nbsp;Michael Weber ,&nbsp;Catherina A.B. Bubb","doi":"10.1016/j.csbj.2024.10.002","DOIUrl":"10.1016/j.csbj.2024.10.002","url":null,"abstract":"<div><div>Continuous, mobile patient monitoring plays a critical role in healthcare, particularly for post-surgery, intermediate care in clinics. The implementation of vital signs monitoring technology enables healthcare professionals to triage patients effectively by maintaining real-time awareness of their health status and allowing for prompt intervention when necessary. This technology supports early mobilization and facilitates the detection of potential complications such as post-surgical sepsis. cosinuss° technology has been evaluated in various studies, in terms of its accuracy in capturing vital parameters and its usability, emphasizing its potential to enhance intermediate patient care and outcomes. This report outlines the design and implementation of cosinuss° Health patient monitoring solution for use in intermediate, postoperative clinic settings. It presents the results and insights from three recent, in-clinic applications, discussing both technical and practical aspects, clinical processes, and the reported satisfaction from both patients and medical caregivers. The findings highlight the promising potential of cosinuss° Health on improving patient monitoring and overall clinical outcomes.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"24 ","pages":"Pages 630-638"},"PeriodicalIF":4.4,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142427953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Text mining approach for feature extraction and cartilage disease grade classification using knee MRI radiology reports","authors":"Antonio Saverio Valente ,&nbsp;Teresa Angela Trunfio ,&nbsp;Marco Aiello ,&nbsp;Dario Baldi ,&nbsp;Marilena Baldi ,&nbsp;Silvio Imbò ,&nbsp;Mario Alessandro Russo ,&nbsp;Carlo Cavaliere ,&nbsp;Monica Franzese","doi":"10.1016/j.csbj.2024.10.003","DOIUrl":"10.1016/j.csbj.2024.10.003","url":null,"abstract":"<div><div>MRI radiology reporting processes can be improved by exploiting structured and semantically labelled data that can be fed to artificial intelligence (AI) tools. AI-based tools assisting radiology reporting can help to automatically individuate cartilage grading in textual magnetic resonance imaging (MRI) reports, thus supporting clinicians' decisions regarding medical imaging utilisation, diagnosis and treatment. In this study, we extracted information (clinical findings, observations, anatomical regions, etc.) and classified knee cartilage degradation from medical reports utilising transfer-learning techniques applied to the Bidirectional Encoder Representations from Transformers (BERT) model and its variants, pre-trained on an Italian-language corpus. To realise this objective, we used a dataset of 750 MRI knee reports written by three radiologists who contributed to a manual annotation process to perform text classification (TC) and named entity recognition (NER) tasks. The dataset was obtained from an internal database of the IRCCS SYNLAB SDN. Seventy percent of the dataset was used for training, 10% was used for validation and 20% was used for testing. The best-performing configurations for NER and TC tasks were based on the pre-trained BERT model. The macro F1-scores obtained with the NER and TC models are 0.89 and 0.81, respectively. The accuracies calculated on the test set for both tasks are 0.96 and 0.99, respectively.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"24 ","pages":"Pages 622-629"},"PeriodicalIF":4.4,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142427955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning and artificial intelligence: Enabling the clinical translation of atomic force microscopy-based biomarkers for cancer diagnosis","authors":"Aidan T. O’Dowling ,&nbsp;Brian J. Rodriguez ,&nbsp;Tom K. Gallagher ,&nbsp;Stephen D. Thorpe","doi":"10.1016/j.csbj.2024.10.006","DOIUrl":"10.1016/j.csbj.2024.10.006","url":null,"abstract":"<div><div>The influence of biomechanics on cell function has become increasingly defined over recent years. Biomechanical changes are known to affect oncogenesis; however, these effects are not yet fully understood. Atomic force microscopy (AFM) is the gold standard method for measuring tissue mechanics on the micro- or nano-scale. Due to its complexity, however, AFM has yet to become integrated in routine clinical diagnosis. Artificial intelligence (AI) and machine learning (ML) have the potential to make AFM more accessible, principally through automation of analysis. In this review, AFM and its use for the assessment of cell and tissue mechanics in cancer is described. Research relating to the application of artificial intelligence and machine learning in the analysis of AFM topography and force spectroscopy of cancer tissue and cells are reviewed. The application of machine learning and artificial intelligence to AFM has the potential to enable the widespread use of nanoscale morphologic and biomechanical features as diagnostic and prognostic biomarkers in cancer treatment.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"24 ","pages":"Pages 661-671"},"PeriodicalIF":4.4,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142526862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"NanoTube Construct: A web tool for the digital construction of nanotubes of single-layer materials and the calculation of their atomistic descriptors powered by Enalos Cloud Platform","authors":"Panagiotis D. Kolokathis ,&nbsp;Dimitrios Zouraris ,&nbsp;Nikolaos K. Sidiropoulos ,&nbsp;Andreas Tsoumanis ,&nbsp;Georgia Melagraki ,&nbsp;Iseult Lynch ,&nbsp;Antreas Afantitis","doi":"10.1016/j.csbj.2024.09.023","DOIUrl":"10.1016/j.csbj.2024.09.023","url":null,"abstract":"<div><div>NanoTube Construct is a web tool for the digital construction of nanotubes based on real and hypothetical single-layer materials including carbon-based materials such as graphene, graphane, graphyne polymorphs, graphidiyene and non-carbon materials such as silicene, germanene, boron nitride, hexagonal bilayer silica, haeckelite silica, molybdene disulfide and tungsten disulfide. Contrary to other available tools, NanoTube Construct has the following features: a) it is not limited to zero thickness materials with specific symmetry, b) it applies energy minimisation to the geometrically constructed Nanotubes to generate realistic ones, c) it derives atomistic descriptors (e.g., the average potential energy per atom, the average coordination number, etc.), d) it provides the primitive unit cell of the constructed Nanotube which corresponds to the selected rolling vector (i.e., the direction in which the starting nanosheet is rolled to form a tube), e) it calculates whether the Nanotube or its corresponding nanosheet is more energetically stable and f) it allows negative chirality indexes. Application of NanoTube Construct for the construction of energy minimised graphane and molybdenum disulfide nanotubes are presented, showcasing the tool's capability. NanoTube Construct is freely accessible through the Enalos Cloud Platform (<span><span>https://enaloscloud.novamechanics.com/diagonal/nanotube/</span><svg><path></path></svg></span>).</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"25 ","pages":"Pages 230-242"},"PeriodicalIF":4.4,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554385","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Performance of a HER2 testing algorithm tailored for urothelial bladder cancer: A Bi-centre study","authors":"Aoling Huang ,&nbsp;Yizhi Zhao ,&nbsp;Feng Guan ,&nbsp;Hongfeng Zhang ,&nbsp;Bin Luo ,&nbsp;Ting Xie ,&nbsp;Shuaijun Chen ,&nbsp;Xinyue Chen ,&nbsp;Shuying Ai ,&nbsp;Xianli Ju ,&nbsp;Honglin Yan ,&nbsp;Lin Yang ,&nbsp;Jingping Yuan","doi":"10.1016/j.csbj.2024.10.007","DOIUrl":"10.1016/j.csbj.2024.10.007","url":null,"abstract":"<div><h3>Aims</h3><div>This study aimed to develop an AI algorithm for automated HER2 scoring in urothelial bladder cancer (UBCa) and assess the interobserver agreement using both manual and AI-assisted methods based on breast cancer criteria.</div></div><div><h3>Methods and Results</h3><div>We utilized 330 slides from two institutions for initial AI development and selected 200 slides for the ring study, involving six pathologists (3 senior, 3 junior). Our AI algorithm achieved high accuracy in two independent tests, with accuracies of 0.94 and 0.92. In the ring study, the AI-assisted method improved both accuracy (0.66 vs 0.94) and consistency (kappa=0.48; 95 % CI, 0.443–0.526 vs kappa=0.87; 95 % CI, 0.852–0.885) compared to manual scoring, especially in HER2-low cases (F1-scores: 0.63 vs 0.92). Additionally, in 62.3 % of heterogeneous HER2-positive cases, the interpretation accuracy significantly improved (0.49 vs 0.93). Pathologists, particularly junior ones, experienced enhanced accuracy and consistency with AI assistance.</div></div><div><h3>Conclusions</h3><div>This is the first study to provide a quantification algorithm for HER2 scoring in UBCa to assist pathologists in diagnosis. The ring study demonstrated that HER2 scoring based on breast cancer criteria can be effectively applied to UBCa. Furthermore, AI assistance significantly improves the accuracy and consistency of interpretations among pathologists with varying levels of experience, even in heterogeneous cases.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"26 ","pages":"Pages 40-50"},"PeriodicalIF":4.4,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142444876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ANN uncertainty estimates in assessing fatty liver content from ultrasound data","authors":"G. Del Corso ,&nbsp;M.A. Pascali ,&nbsp;C. Caudai ,&nbsp;L. De Rosa ,&nbsp;A. Salvati ,&nbsp;M. Mancini ,&nbsp;L. Ghiadoni ,&nbsp;F. Bonino ,&nbsp;M.R. Brunetto ,&nbsp;S. Colantonio ,&nbsp;F. Faita","doi":"10.1016/j.csbj.2024.09.021","DOIUrl":"10.1016/j.csbj.2024.09.021","url":null,"abstract":"<div><h3>Background and objective</h3><div>This article uses three different probabilistic convolutional architectures applied to ultrasound image analysis for grading Fatty Liver Content (FLC) in Metabolic Dysfunction Associated Steatotic Liver Disease (MASLD) patients. Steatosis is a new silent epidemic and its accurate measurement is an impelling clinical need, not only for hepatologists, but also for experts in metabolic and cardiovascular diseases. This paper aims to provide a robust comparison between different uncertainty quantification strategies to identify advantages and drawbacks in a real clinical setting.</div></div><div><h3>Methods</h3><div>We used a classical Convolutional Neural Network, a Monte Carlo Dropout, and a Bayesian Convolutional Neural Network with the goal of not only comparing the goodness of the predictions, but also to have access to an evaluation of the uncertainty associated with the outputs.</div></div><div><h3>Results</h3><div>We found that even if the prediction based on a single ultrasound view is reliable (relative RMSE [5.93%-12.04%]), networks based on two ultrasound views outperform them (relative RMSE [5.35%-5.87%]). In addition, the results show that the introduction of a “not confident” category contributes to increase the percentage of correctly predicted cases and to decrease the percentage of mispredicted cases, especially for semi-intrusive methods.</div></div><div><h3>Conclusions</h3><div>The possibility of having access to information about the confidence with which the network produces its outputs is a great advantage, both from the point of view of physicians who want to use neural networks as computer-aided diagnosis, and for developers who want to limit overfitting and obtain information about dataset problems in terms of out-of-distribution detection.</div></div>","PeriodicalId":10715,"journal":{"name":"Computational and structural biotechnology journal","volume":"24 ","pages":"Pages 603-610"},"PeriodicalIF":4.4,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142427954","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}