Examinando por Autor "Zahia, Sofia"

Mostrando 1 - 6 de 6
  • Miniatura
    Ítem
    Breast cancer histopathology image classification using an ensemble of deep learning models
    (MDPI AG, 2020-08-02) Hameed, Zabit; Zahia, Sofia; García-Zapirain, Begoña; Aguirre, José Javier; Vanegas, Ana María
    Breast cancer is one of the major public health issues and is considered a leading cause of cancer-related deaths among women worldwide. Its early diagnosis can effectively help in increasing the chances of survival rate. To this end, biopsy is usually followed as a gold standard approach in which tissues are collected for microscopic analysis. However, the histopathological analysis of breast cancer is non-trivial, labor-intensive, and may lead to a high degree of disagreement among pathologists. Therefore, an automatic diagnostic system could assist pathologists to improve the effectiveness of diagnostic processes. This paper presents an ensemble deep learning approach for the definite classification of non-carcinoma and carcinoma breast cancer histopathology images using our collected dataset. We trained four different models based on pre-trained VGG16 and VGG19 architectures. Initially, we followed 5-fold cross-validation operations on all the individual models, namely, fully-trained VGG16, fine-tuned VGG16, fully-trained VGG19, and fine-tuned VGG19 models. Then, we followed an ensemble strategy by taking the average of predicted probabilities and found that the ensemble of fine-tuned VGG16 and fine-tuned VGG19 performed competitive classification performance, especially on the carcinoma class. The ensemble of fine-tuned VGG16 and VGG19 models offered sensitivity of 97.73% for carcinoma class and overall accuracy of 95.29%. Also, it offered an F1 score of 95.29%. These experimental results demonstrated that our proposed deep learning approach is effective for the automatic classification of complex-natured histopathology images of breast cancer, more specifically for carcinoma images.
  • Miniatura
    Ítem
    A comparative study between scanning devices for 3D printing of personalized ostomy patches
    (MDPI, 2022-01-12) Zahia, Sofia; García-Zapirain, Begoña; Anakabe, Jon ; Páez Orue-Echevarría, Joan Ander; Bastidas-Jossa, Óscar; Loizate Totoricagüena, Alberto
    This papers presents a comparative study of three different 3D scanning modalities to acquire 3D meshes of stoma barrier rings from ostomized patients. Computerized Tomography and Structured light scanning methods were the digitization technologies studied in this research. Among the Structured Light systems, the Go!Scan 20 and the Structure Sensor were chosen as the handheld 3D scanners. Nineteen ostomized patients took part in this study, starting from the 3D scans acquisition until the printed ostomy patches validation. 3D mesh processing, mesh generation and 3D mesh comparison was carried out using commercial softwares. The results of the presented study show that the Structure Sensor, which is the low cost structured light 3D sensor, has a great potential for such applications. This study also discusses the benefits and reliability of low-cost structured light systems.
  • Miniatura
    Ítem
    Dyslexia detection using 3D convolutional neural networks and functional magnetic resonance imaging
    (Elsevier Ireland Ltd, 2020-12) Zahia, Sofia; García-Zapirain, Begoña; Saralegui, Ibone; Fernandez-Ruanova, Begoña
    Background and Objectives: Dyslexia is a disorder of neurological origin which affects the learning of those who suffer from it, mainly children, and causes difficulty in reading and writing. When undiagnosed, dyslexia leads to intimidation and frustration of the affected children and also of their family circles. In case no early intervention is given, children may reach high school with serious achievement gaps. Hence, early detection and intervention services for dyslexic students are highly important and recommended in order to support children in developing a positive self-esteem and reaching their maximum academic capacities. This paper presents a new approach for automatic recognition of children with dyslexia using functional magnetic resonance Imaging. Methods: Our proposed system is composed of a sequence of preprocessing steps to retrieve the brain activation areas during three different reading tasks. Conversion to Nifti volumes, adjustment of head motion, normalization and smoothing transformations were performed on the fMRI scans in order to bring all the subject brains into one single model which will enable voxels comparison between each subject. Subsequently, using Statistical Parametric Maps (SPMs), a total of 165 3D volumes containing brain activation of 55 children were created. The classification of these volumes was handled using three parallel 3D Convolutional Neural Network (3D CNN), each corresponding to a brain activation during one reading task, and concatenated in the last two dense layers, forming a single architecture devoted to performing optimized detection of dyslexic brain activation. Additionally, we used 4-fold cross validation method in order to assess the generalizability of our model and control overfitting. Results: Our approach has achieved an overall average classification accuracy of 72.73%, sensitivity of 75%, specificity of 71.43%, precision of 60% and an F1-score of 67% in dyslexia detection. Conclusions: The proposed system has demonstrated that the recognition of dyslexic children is feasible using deep learning and functional magnetic resonance Imaging when performing phonological and orthographic reading tasks.
  • Miniatura
    Ítem
    Integrating 3D model representation for an accurate non-invasive assessment of pressure injuries with deep learning
    (MDPI AG, 2020-05-21) Zahia, Sofia; García-Zapirain, Begoña; Elmaghraby, Adel Said
    Pressure injuries represent a major concern in many nations. These wounds result from prolonged pressure on the skin, which mainly occur among elderly and disabled patients. If retrieving quantitative information using invasive methods is the most used method, it causes significant pain and discomfort to the patients and may also increase the risk of infections. Hence, developing non-intrusive methods for the assessment of pressure injuries would represent a highly useful tool for caregivers and a relief for patients. Traditional methods rely on findings retrieved solely from 2D images. Thus, bypassing the 3D information deriving from the deep and irregular shape of this type of wounds leads to biased measurements. In this paper, we propose an end-to-end system which uses a single 2D image and a 3D mesh of the pressure injury, acquired using the Structure Sensor, and outputs all the necessary findings such as: external segmentation of the wound as well as its real-world measurements (depth, area, volume, major axis and minor axis). More specifically, a first block composed of a Mask RCNN model uses the 2D image to output the segmentation of the external boundaries of the wound. Then, a second block matches the 2D and 3D views to segment the wound in the 3D mesh using the segmentation output and generates the aforementioned real-world measurements. Experimental results showed that the proposed framework can not only output refined segmentation with 87% precision, but also retrieves reliable measurements, which can be used for medical assessment and healing evaluation of pressure injuries.
  • Miniatura
    Ítem
    Predicting physical exercise adherence in fitness apps using a deep learning approach
    (MDPI, 2021-10-14) Bastidas-Jossa, Óscar; Zahia, Sofia; Fuente Vidal, Andrea; Sánchez Férez, Néstor; Roda Noguera, Oriol; Montane, Joel; García-Zapirain, Begoña
    The use of mobile fitness apps has been on the rise for the last decade and especially during the worldwide SARS-CoV-2 pandemic, which led to the closure of gyms and to reduced outdoor mobility. Fitness apps constitute a promising means for promoting more active lifestyles, although their attrition rates are remarkable and adherence to their training plans remains a challenge for developers. The aim of this project was to design an automatic classification of users into adherent and non-adherent, based on their training behavior in the first three months of app usage, for which purpose we proposed an ensemble of regression models to predict their behaviour (adherence) in the fourth month. The study was conducted using data from a total of 246 Mammoth Hunters Fitness app users. Firstly, pre-processing and clustering steps were taken in order to prepare the data and to categorize users into similar groups, taking into account the first 90 days of workout sessions. Then, an ensemble approach for regression models was used to predict user training behaviour during the fourth month, which were trained with users belonging to the same cluster. This was used to reach a conclusion regarding their adherence status, via an approach that combined affinity propagation (AP) clustering algorithm, followed by the long short-term memory (LSTM), rendering the best results (87% accuracy and 85% F1_score). This study illustrates the suggested the capacity of the system to anticipate future adherence or non-adherence, potentially opening the door to fitness app creators to pursue advanced measures aimed at reducing app attrition.
  • No hay miniatura disponible
    Ítem
    Pressure injury image analysis with machine learning techniques: a systematic review on previous and possible future methods
    (Elsevier B.V., 2020-01) Zahia, Sofia; García-Zapirain, Begoña; Sevillano, Xavier; González, Alejandro; Kim, Paul J.; Elmaghraby, Adel Said
    Pressure injuries represent a tremendous healthcare challenge in many nations. Elderly and disabled people are the most affected by this fast growing disease. Hence, an accurate diagnosis of pressure injuries is paramount for efficient treatment. The characteristics of these wounds are crucial indicators for the progress of the healing. While invasive methods to retrieve information are not only painful to the patients but may also increase the risk of infections, non-invasive techniques by means of imaging systems provide a better monitoring of the wound healing processes without causing any harm to the patients. These systems should include an accurate segmentation of the wound, the classification of its tissue types, the metrics including the diameter, area and volume, as well as the healing evaluation. Therefore, the aim of this survey is to provide the reader with an overview of imaging techniques for the analysis and monitoring of pressure injuries as an aid to their diagnosis, and proof of the efficiency of Deep Learning to overcome this problem and even outperform the previous methods. In this paper, 114 out of 199 papers retrieved from 8 databases have been analyzed, including also contributions on chronic wounds and skin lesions.