Browsing by Author "Coelho, Luis"
Now showing 1 - 10 of 31
Results Per Page
Sort Options
- Adaptive modeling and high quality spectral estimation for speech enhancementPublication . Coelho, Luis; Braga, DanielaIn this work an adaptive modeling and spectral estimation scheme based on a dual Discrete Kalman Filtering (DKF) is proposed for speech enhancement. Both speech and noise signals are modeled by an autoregressive structure which provides an underlying time frame dependency and improves time-frequency resolution. The model parameters are arranged to obtain a combined state-space model and are also used to calculate instantaneous power spectral density estimates. The speech enhancement is performed by a dual discrete Kalman filter that simultaneously gives estimates for the models and the signals. This approach is particularly useful as a pre-processing module for parametric based speech recognition systems that rely on spectral time dependent models. The system performance has been evaluated by a set of human listeners and by spectral distances. In both cases the use of this pre-processing module has led to improved results.
- An automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preferencePublication . Coelho, Luis; Braga, Daniela; Sales-Dias, Miguel; Garcia-Mateo, CarmenIn the last few years the number of systems and devices that use voice based interaction has grown significantly. For a continued use of these systems the interface must be reliable and pleasant in order to provide an optimal user experience. However there are currently very few studies that try to evaluate how good is a voice when the application is a speech based interface. In this paper we present a new automatic voice pleasantness classification system based on prosodic and acoustic patterns of voice preference. Our study is based on a multi-language database composed by female voices. In the objective performance evaluation the system achieved a 7.3% error rate.
- Análise Objetiva ao Desempenho do Monofilamento de Semmes-Weinstein de 10 gf para Avaliação da Sensibilidade Cutânea Protetora no Pé DiabéticoPublication . Martins, Pedro; Coelho, LuisA perda de sensibilidade protetora no pé diabético obriga a um diagnóstico precoce devido à eminente possibilidade de ulceração e amputação do membro afetado. Na avaliação da perda de sensibilidade protetora, o monofilamento de Semmes-Weinstein (MSW) de 10 gf é o instrumento de primeira linha mais utilizado. Este estudo foi conduzido com o objetivo de realizar uma análise objetiva e metrológica ao desempenho do MSW de 10 gf, recorrendo a um equipamento de medição previamente desenvolvido para o efeito. Procedeu-se a uma verificação metrológica de três MSW de 10 gf através de um método automatizado sob condições controladas. A força de compressão dos MSW revelou-se altamente dependente do número de utilizações, da temperatura ambiente e da velocidade de deslocamento e contacto com a região cutânea a avaliar. Efetivamente todos os MSW testados manifestaram dependência de fatores externos na sua força de compressão e registaram erros relativos superiores a 10 %, valor considerado crítico para discriminar uma possível evolução da patologia.
- Automatic Classification and Segmentation of Low-Grade Gliomas in Magnetic Resonance ImagingPublication . Barbosa, Marta; Moreira, Pedro; Ribeiro, Rogério; Coelho, LuisIn this article a new methodology is proposed to tackle the problem of automatic segmentation of low-grade gliomas. The possibility of knowing the limits of this type of tumor is crucial for effectively characterizing the neoplasm, enabling, in certain cases, to obtain useful information about how to treat the patient in a more effective way. Using a database of magnetic resonance images, containing several occurrences of this type of tumors, and through a carefully designed image processing pipeline, the purpose of this work is to accurately locate, isolate and thus facilitate the classification of the pathology. The proposed methodology, described in detail, was able to achieve an accuracy of 87.5% for a binary classification task. The quality of the identified regions had an accuracy of 81.6%. These are promising results that may point the effectiveness of the approach. The low contrast of the images, as a result of the acquisition process, and the detection of very small tumors are still challenges that bring motivation to further pursue additional results.
- Automatic Segmentation of Monofilament Testing Sites in Plantar Images for Diabetic Foot ManagementPublication . Costa, Tatiana; Coelho, Luis; Silva, Manuel F.Diabetic peripheral neuropathy is a major complication of diabetes mellitus, and it is the leading cause of foot ulceration and amputations. The Semmes–Weinstein monofilament examination (SWME) is a widely used, low-cost, evidence-based tool for predicting the prognosis of diabetic foot patients. The examination can be quick, but due to the high prevalence of the disease, many healthcare professionals can be assigned to this task several days per month. In an ongoing project, it is our objective to minimize the intervention of humans in the SWME by using an automated testing system relying on computer vision. In this paper we present the project’s first part, constituting a system for automatically identifying the SWME testing sites from digital images. For this, we have created a database of plantar images and developed a segmentation system, based on image processing and deep learning—both of which are novelties. From the 9 testing sites, the system was able to correctly identify most 8 in more than 80% of the images, and 3 of the testing sites were correctly identified in more than 97.8% of the images.
- Benchmarking computer-vision-based facial emotion classification algorithms while wearing surgical masksPublication . Coelho, Luis; Reis, Sara; Moreira, Cristina; Cardoso, Helena; Sequeira, Miguela; Coelho, RaquelEffective human communication relies heavily on emotions, making them a crucial aspect of interaction. As technology progresses, the desire for machines to exhibit more human-like characteristics, including emotion recognition, grows. DeepFace has emerged as a widely adopted library for facial emotion recognition. However, the widespread use of surgical masks after the COVID-19 pandemic presents a considerable obstacle to its performance. To assess this issue, we conducted a benchmark using the FER2013 dataset. The results revealed a substantial performance decline when individuals wore surgical masks. “Disgust” suffers a 22.6% F1-score reduction, while “Surprise” is least affected with a 48.7% reduction. Addressing these issues improves human–machine interfaces and paves the way for more natural machine communication.
- Biomedical Students’ Motivation with Project Based Learning: A Case StudyPublication . Reis, Sara Seabra dos; Coelho, Maria De Fátima; Coelho, LuisOne of the teacher’s first goals should be to motivate students to learn. Involving students in the learning process using project-based learning could be a useful and powerful tool to prepare the students for them professional future. As part of a degree course in Biomedical Engineering, from a Portuguese University, students were asked to look at society and identify a possible failure related to the biomedical field. From the identification of this fault, the project consisted in the planning and its possible resolution. In this article we present the case study of a student’s team, whose project was candidate and winner for a national prize in the field of health innovation. Despite the particularization of this case study, the students considered the experience innovative and simultaneously motivating. They also highlighted the added value of a project going beyond the classroom. Therefore, in addition to allowing them to obtain classification to the curricular unit, this type of teaching and learning could have a special impact on civil society.
- Calibration and Modeling of the Semmes–Weinstein Monofilament for Diabetic Foot ManagementPublication . Martins, Pedro; Coelho, Luis; Campilho, Raul Duarte Salgueiral GomesDiabetic foot is a serious complication that poses significant risks for diabetic patients. The resulting reduction in protective sensitivity in the plantar region requires early detection to prevent ulceration and ultimately amputation. The primary method employed for evaluating this sensitivity loss is the 10 gf Semmes–Weinstein monofilament test, commonly used as a first-line procedure. However, the lack of calibration in existing devices often introduces decision errors due to unreliable feedback. In this article, the mechanical behavior of a monofilament was analytically modeled, seeking to promote awareness of the impact of different factors on clinical decisions. Furthermore, a new device for the automation of the metrological evaluation of the monofilament is described. Specific testing methodologies, used for the proposed equipment, are also described, creating a solid base for the establishment of future calibration guidelines. The obtained results showed that the tested monofilaments had a very high error compared to the 10 gf declared by the manufacturers. To improve the precision and reliability of assessing the sensitivity loss, the frequent metrological calibration of the monofilament is crucial. The integration of automated verification, simulation capabilities, and precise measurements shows great promise for diabetic patients, reducing the likelihood of adverse outcomes.
- CardioML: integrating personal cardiac information for ubiquous diagnosis and analysisPublication . Coelho, Luis; Queirós, RicardoThe latest medical diagnosis devices enable the performance of e-diagnosis making the access to these services easier, faster and available in remote areas. However this imposes new communications and data interchange challenges. In this paper a new XML based format for storing cardiac signals and related information is presented. The proposed structure encompasses data acquisition devices, patient information, data description, pathological diagnosis and waveform annotation. When compared with similar purpose formats several advantages arise. Besides the full integrated data model it may also be noted the available geographical references for e-diagnosis, the multi stream data description, the ability to handle several simultaneous devices, the possibility of independent waveform annotation and a HL7 compliant structure for common contents. These features represent an enhanced integration with existent systems and an improved flexibility for cardiac data representation.
- Emerging advancements for virtual and augmented reality in healthcarePublication . Coelho, Luis; Queirós, Ricardo; Reis, Sara SeabraWithin the last few years, devices that are increasingly capable of offering an immersive experience close to reality have emerged. As devices decrease in size, the interest and application possibilities for them increase. In the healthcare sector, there is an enormous potential for virtual reality development, as this technology allows, on the one hand, the execution of operations or processes at a distance, decoupling realities; and on the other hand, it offers the possibility of simulation for training purposes, whenever there are contexts of risk to the patient or to the health professional. However, virtual reality devices and immersion in virtual environments still requires some improvement as complaints such as headaches and nausea are still common among users, and so continuous research and development is critical to progress the technology. Emerging Advancements for Virtual and Augmented Reality in Healthcare synthesizes the trends, best practices, methodologies, languages, and tools used to implement virtual reality and create a positive user experience while also discussing how to implement virtual reality into day-to-day work with a focus on healthcare professionals and related areas. The application possibilities and their impact are transversal to all areas of health and fields such as education, training, surgery, pain management, physical rehabilitation, stroke rehabilitation, phobia therapy, and telemedicine. Covering topics such as mental health treatment and virtual simulations, it is ideal for medical professionals, engineers, computer scientists, researchers, practitioners, managers, academicians, teachers, and students.