Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Contribución de congreso (ConferenceOutput)

Deep Learning Based Non-invasive Meningitis Screening Using High-Resolution Ultrasound in Neonates and Infants from Mozambique, Spain and Morocco

  • Jobst, Beatrice M.
  • Carandell, Francesc
  • Ajanovic, Sara
  • Sial, Hassan
  • Jiménez, Javier
  • Quesada, Rita
  • Santos, Fabião
  • Lopez-Azorín, Manuela
  • Valverde, Eva
  • Ybarra, Marta
  • Bravo, M. Carmen
  • Muñoz, David
  • Agut, Thais
  • Salas, Barbara
  • Carreras, Nuria
  • Alarcón, Ana
  • Iriondo, Martín
  • Luaces, Carles
  • Sidat, Muhammad
  • Zandamela, Mastalina
  • Rodrigues, Paula
  • Martins, Luzidina
  • Cassia, Uneisse
  • Bramugy, Justina
  • Cossa, Anelsio
  • Mucasse, Campos
  • Buck, W. Chris
  • Arias, Sara
  • El Abbass, Chaymae
  • Tligui, Houssain
  • Barkat, Amina
  • Amalik, Najat
  • Zizi, Imane
  • Ibáñez Rodríguez, Alberto
  • Parrilla, Montserrat
  • Elvira Segura, Luis
  • Calvo, Cristina
  • Pellicer, Adelina
  • Cabañas, Fernando
  • Bassat, Quique
  • Petrone, Paula
1st MICCAI Meets Africa Workshop, MImA 2024 and 1st MICCAI Student Board Workshop on Empowering Medical Information Computing and Research through Early-Career Expertise, EMERGE 2024, Held in Conjunction with MICCAI 2024. Medical Information Computing. MImA EMERGE 2024 2024. Communications in Computer and Information Science, vol 2240., Meningitis is a life-threatening disease, resulting in severe neurological damage or death if not treated in time. The standard diagnostic method is an invasive lumbar puncture (LP), not exempt of complications and not always feasible, especially in resource-poor settings. To overcome these challenges, we developed Neosonics®, a novel non-invasive high-resolution ultrasound (HRUS) technology that detects backscatter signals from white blood cells (WBCs) in cerebrospinal fluid (CSF) below the infant fontanel. Using deep learning (DL) for image analysis, this technology classifies patients based on WBC levels, providing a rapid, non-invasive screening method for infant meningitis and with this, limiting indications for LPs for positive cases only, if not contraindicated. A convolutional neural network was trained and validated using CSF HRUS patient data gathered from cohorts from three different countries, Mozambique, Spain and Morocco. Then, a soft voting ensemble-learning technique was applied to give a classification on a patient-level. The DL model showed on a single-image level a sensitivity (SE) and specificity (SP) of 71.1% and 87.0% for the Mozambican cohort, SE = 73.3% and SP = 75.5% for the Spanish cohort, and SE = 72.0% and SP = 88.7% for the Moroccan cohort. On a patient-level, the overall SE and SP was 94.4% and 94.8%, respectively. We demonstrate in this study the significance of combining HRUS and DL for the non-invasive detection of infant meningitis, offering an efficient, cost-effective solution suitable also for limited resource settings and remote areas., Kriba acknowledges support from the European Union’s Horizon Europe research and innovation programme, project code 190155553 - NEOSONICS. All authors with ISGlobal affiliation acknowledge support from the grant CEX2018-000806-S funded by MCIN/AEI/ https://doi.org/10.13039/501100011033, and support from the Generalitat de Catalunya through the CERCA Program. Mozambique: This work was supported by the Bill & Melinda Gates Foundation INV-048197. Morocco: This project was supported in part by Spanish Agency of Cooperation (AECID) under the reference number 2019/ACDE/001196. Project PI16/01822 from PI ENRIQUE BASSAT ORELLANA and project PI16/00738 funded by Instituto de Salud Carlos III, co-funded by the European Union (FEDER) “Una manera de hacer Europa”., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Estimation of the concentration of particles in suspension based on envelope statistics of ultrasound backscattering

  • Fernández, Alba
  • Ibáñez Rodríguez, Alberto
  • Parrilla Romero, Montserrat
  • Elvira Segura, Luis
  • Bassat, Quique
  • Jiménez, Javier
13 páginas, 17 figuras, 2 tablas, This work deals with the development of a methodology to evaluate the concentration in cell or particle sus- pensions from ultrasound images. The novelty of the method is based on two goals: first, it should be valid when the energy reaching the scatterers is unknown and cannot be measured or calibrated. In addition, it should be robust against echo overlap which may occur due to high scatterer concentration. Both characteristics are especially valuable in quantitative ultrasound analysis in the clinical context. In this regard, the present work considers the ability of envelope statistics models to characterize ultrasound images. Envelope statistical analysis are based on the examination of the physical properties of a medium through the study of the statistical dis- tribution of the backscattered signal envelop. A review of the statistical distributions typically used to charac- terize scattering mediums was conducted. The main parameters of the distribution were estimated from simulations of signals backscattered by particle suspensions. Then, the ability of these parameters to characterize the suspension concentration was analyzed and the μ parameter from the Homodyned-K distribution resulted as the most suitable parameter for the task. Simulations were also used to study the impact of noise, signal amplitude variability and dispersion of particle sizes on the estimation method. The efficiency of the algorithm on experimental measurements was also evaluated. To this end, two sets of ultrasound images were obtained from suspensions of 7 μm and 12 μm polystyrene particles in water, using a 20 MHz focused transducer. The methodology proved to be efficient to quantify the concentration of particle suspensions in the range between 5 and 3000 particles/μl, achieving similar results for both particle sizes and for different signal-to-noise ratios., This work was supported by the Instituto de Salud Carlos III, projects PI16/00738 and PI16/01822, and co-financed by FEDER resources, by the project PID2019-111392RB-I00 (Spanish Ministry of Science and Innovation) and by the contract GARJUV_CAM_18_00217 from the Youth Employment Program (Madrid Department of Education and Research), co-financed by the European Social Fund (ESF). ISGlobal also receives support from the Spanish Ministry of Science and Innovation through the “Centro de Excelencia Severo Ochoa 2019-2023” Program (CEX2018-000806-S), and support from the Generalitat de Catalunya through the CERCA Program. CISM is supported by the Government of Mozambique and the Spanish Agency for International Development (AECID), Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

A New Methodology for the Assessment of Very Low Concentrations of Cells in Serous Body Fluids Based on the Count of Ultrasound Echoes Backscattered From Cells

  • Elvira Segura, Luis
  • Ibáñez Rodríguez, Alberto
  • Fernández, Alba
  • Durán Gómez, Carmen
  • Parrilla Romero, Montserrat
  • Pose Díez de la Lastra, Alicia
  • Bassat, Quique
  • Jiménez, Javier
13 páginas, 10 figuras, Amethodology for the assessmentof cell concentration, in the range 5–100 cells/μL, suitable for in vivo analysis of serous body fluids is presented in this work. This methodology is based on the quantitative analysis of ultrasound images obtained fromcell suspensionsand considers applicability criteria, such as short analysis times, moderate frequency, and absolute concentration estimation, all necessary to deal with the variability of tissues among different patients. Numerical simulations provided the framework to analyze the impact of echo overlapping and the polydispersion of scatterer sizes on the cell concentration estimation. The cell concentration range that can be analyzed as a function of the transducer and emitted waveformused was also discussed. Experimentswere conducted to evaluate the performance of the method using 7-μmand 12-μmpolystyreneparticles inwater suspensions in the 5–100 particles/μL range. A single scanning focused transducer working at a central frequency of 20 MHz was used to obtain ultrasound images. The method proposed to estimate the concentration proved to be robust for different particle sizes and variations of gain acquisition settings. The effect of tissues placed in the ultrasound path between the probe and the sample was also investigated using 3-mm-thick tissuemimics. Under this situation, the algorithm was robust for the concentration analysis of 12 μm particle suspensions, yet significant deviations were obtained for the smallest particles., This work was supported in part by the Instituto de Salud Carlos III under Project PI16/00738 and PI16/01822, in part by FEDER (European Regional Development Fund) resources, and in part by the Spanish Ministry of Science and Innovation under Project PID2019-111392RB-I00. The work of Alba Fernández was supported in part by the Comunidad de Madrid and Alicia Pose through the JAEINTRO Programme under Contract PEJD-2018-PRE/IND-9090 and in part by the Consejo Superior de Investigaciones Científicas (CSIC) under Grant JAEINT_18_01152. The work of Quique Bassat was supported in part by the Spanish Ministry of Science and Innovation through the Centro de Excelencia Severo Ochoa 2019-2023 Program under Grant CEX2018-000806-S, in part by the Generalitat de Catalunya through the CERCA Program, in part by the Government of Mozambique, and in part by the Spanish Agency for International Development (AECID)., Peer reviewed