Gemma Urbanos
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Gemma Urbanos has a total of 48 co-authors. Here are the top co-authors by number of shared publications:
- Alfonso Lagares
Department of Neurosurgery, Hospital Universitario 12 de Octubre, Madrid, Spain
6 shared publications - Cesar Sanz
Universidad Politécnica de Madrid, Spain
6 shared publications - Eduardo Juarez
Universidad Politécnica de Madrid, Spain
6 shared publications - Alberto Martín-Pérez
CEIMM, Center for Industrial Electronics and Multimodal Systems, Universidad Politécnica de Madrid, 28031 Madrid, Spain
5 shared publications - Guillermo Vazquez
Research Center on Software Technologies and Multimedia Systems, Universidad Politécnica de Madrid (UPM), 28031 Madrid, Spain
5 shared publications - Manuel Villa
Research Center on Software Technologies and Multimedia Systems, Universidad Politécnica de Madrid, 28031 Madrid, Spain
5 shared publications
Latest Publications
- 23 April 2026
- Published by Universidad Politecnica de Madrid - University Library
Contribuciones a la caracterización automática del tejido cerebral y lesiones asociadas mediante modelos supervisados en imagen médica multimodal para diagnóstico asistido
Clinical practice in neuroscience continues to face persistent limitations in diagnostic accuracy, reproducibility, and efficiency. Visual interpretation of medical imaging is inherently subjective, clinical scales provide only a partial characterization of the underlying pathophysiology, and manual lesion segmentation is time- and resource-intensive. These limitations are particularly critical in acute conditions such as aneurysmal subarachnoid hemorrhage (aSAH), where delays in lesion quantification may adversely affect prognosis and therapeutic management. In this context, this doctoral thesis develops and validates automated medical image analysis systems based on machine learning (ML) to improve the objectivity, reproducibility, and efficiency of radiological analysis, as well as the prediction of clinical outcomes and decision support in neurosurgery. The work is structured around three main axes: (i) intraoperative tissue classification using hyperspectral imaging (HSI) in glioblastoma multiforme (GBM), (ii) automatic segmentation of intracranial hemorrhages on computed tomography (CT) in aSAH, and (iii) prognostic modeling based on radiomic biomarkers derived from automated segmentations of hemorrhages and surrounding brain parenchyma. The first axis was developed in a prospective cohort of 12 patients with GBM treated at Hospital Universitario 12 de Octubre, with acquisition of 13 HSI cubes using a snapshot system with 25 spectral bands (655-975 nm). Five tissue classes were segmented and ML models were trained for tissue classification. Intra-patient validation achieved overall accuracies above 95%, while inter-patient evaluation (leave-one-patient-out) yielded an overall accuracy of 60%, highlighting the impact of biological variability. Band-wise efficiency outperformed previous studies with higher spectral resolution, and the resulting data were incorporated into the public SLIMBRAIN database. The second axis addresses automatic hemorrhage segmentation on CT in an institutional cohort of 539 patients with aSAH, of whom 223 had manual segmentations, augmented with 450 publicly available studies. Using the nnU-Net framework, five-fold cross-validation, and model ensembling, the best-performing model achieved a median Dice coefficient of 0.81, outperforming previous methods both quantitatively and qualitatively according to neurosurgeon evaluation. Automatic segmentation reduced analysis time from approximately 15 minutes to 23.4 seconds, and the derived hemorrhage volume demonstrated prognostic performance comparable to manual volumetry and superior to the modified Fisher clinical scale. The third axis evaluates the prediction of mortality, functional outcome, vasospasm, and hydrocephalus in 444 patients with aSAH. Models based on clinical variables, radiomic descriptors, and their combination were developed using nested cross-validation, achieving mean areas under the curve of up to 0.89 and consistently outperforming established clinical scales. Interpretability analysis highlight clinical variables such as the World Federation of Neurosurgical Societies (WFNS) scale and radiomic descriptors of the brain parenchyma as key predictive factors. Additionally, a radiomic model was developed for the automatic estimation of the SEBES (Subarachnoid Hemorrhage Early Brain Edema Score), showing high concordance with expert assessment, albeit with lower prognostic value compared with models specifically optimized for each clinical outcome. Overall, this thesis demonstrates that the integration of ML algorithms with biomarkers derived from HSI and radiomics enables a more objective, reproducible, and efficient quantitative analysis of neurological pathology, although clinical generalization will require prospective multicenter validation. RESUMEN La práctica clínica en neurociencia se enfrenta a limitaciones persistentes en precisión diagnóstica, reproducibilidad y eficiencia. La interpretación visual de la imagen médica es subjetiva, las escalas clínicas ofrecen una caracterización fisiopatológica parcial y la segmentación manual de lesiones es costosa en tiempo y recursos. Estas limitaciones son especialmente críticas en patologías agudas como la hemorragia subaracnoidea aneurismática (HSAa), donde los retrasos en la cuantificación pueden afectar al pronóstico y al manejo terapéutico. En este contexto, esta tesis doctoral desarrolla y valida sistemas automatizados de análisis de imagen médica basados en aprendizaje automático (Machine Learning, ML) para mejorar la objetividad, reproducibilidad y eficiencia del análisis radiológico, así como la predicción de desenlaces clínicos y el apoyo a la toma de decisiones en neurocirugía. El trabajo se articula en tres ejes: (i) clasificación tisular intraoperatoria mediante imagen hiperespectral (HSI) en glioblastoma multiforme (GBM), (ii) segmentación automática de hemorragias intracraneales en tomografía computarizada (TC) en HSAa y (iii) modelización pronóstica basada en biomarcadores radiómicos derivados de segmentaciones automáticas de las hemorragias y del parénquima cerebral adyacente. El primer eje se desarrolla en una cohorte prospectiva de 12 pacientes con GBM intervenidos en el Hospital Universitario 12 de Octubre, con adquisición de 13 cubos HSI mediante un sistema snapshot de 25 bandas (655-975 nm). Se han segmentado cinco clases tisulares y se han entrenado modelos de ML para clasificación tisular. La validación intrapaciente alcanza precisiones globales mayores de 95%, mientras que en el escenario interpaciente (leave-one-patient-out) la precisión global es del 60%, evidenciando la variabilidad biológica. La eficiencia por banda supera la de estudios previos con mayor resolución espectral y los datos se incorporaron a la base...
- 23 April 2026
- Published by Universidad Politecnica de Madrid - University Library
Contribuciones a la caracterización automática del tejido cerebral y lesiones asociadas mediante modelos supervisados en imagen médica multimodal para diagnóstico asistido
Clinical practice in neuroscience continues to face persistent limitations in diagnostic accuracy, reproducibility, and efficiency. Visual interpretation of medical imaging is inherently subjective, clinical scales provide only a partial characterization of the underlying pathophysiology, and manual lesion segmentation is time- and resource-intensive. These limitations are particularly critical in acute conditions such as aneurysmal subarachnoid hemorrhage (aSAH), where delays in lesion quantification may adversely affect prognosis and therapeutic management. In this context, this doctoral thesis develops and validates automated medical image analysis systems based on machine learning (ML) to improve the objectivity, reproducibility, and efficiency of radiological analysis, as well as the prediction of clinical outcomes and decision support in neurosurgery. The work is structured around three main axes: (i) intraoperative tissue classification using hyperspectral imaging (HSI) in glioblastoma multiforme (GBM), (ii) automatic segmentation of intracranial hemorrhages on computed tomography (CT) in aSAH, and (iii) prognostic modeling based on radiomic biomarkers derived from automated segmentations of hemorrhages and surrounding brain parenchyma. The first axis was developed in a prospective cohort of 12 patients with GBM treated at Hospital Universitario 12 de Octubre, with acquisition of 13 HSI cubes using a snapshot system with 25 spectral bands (655-975 nm). Five tissue classes were segmented and ML models were trained for tissue classification. Intra-patient validation achieved overall accuracies above 95%, while inter-patient evaluation (leave-one-patient-out) yielded an overall accuracy of 60%, highlighting the impact of biological variability. Band-wise efficiency outperformed previous studies with higher spectral resolution, and the resulting data were incorporated into the public SLIMBRAIN database. The second axis addresses automatic hemorrhage segmentation on CT in an institutional cohort of 539 patients with aSAH, of whom 223 had manual segmentations, augmented with 450 publicly available studies. Using the nnU-Net framework, five-fold cross-validation, and model ensembling, the best-performing model achieved a median Dice coefficient of 0.81, outperforming previous methods both quantitatively and qualitatively according to neurosurgeon evaluation. Automatic segmentation reduced analysis time from approximately 15 minutes to 23.4 seconds, and the derived hemorrhage volume demonstrated prognostic performance comparable to manual volumetry and superior to the modified Fisher clinical scale. The third axis evaluates the prediction of mortality, functional outcome, vasospasm, and hydrocephalus in 444 patients with aSAH. Models based on clinical variables, radiomic descriptors, and their combination were developed using nested cross-validation, achieving mean areas under the curve of up to 0.89 and consistently outperforming established clinical scales. Interpretability analysis highlight clinical variables such as the World Federation of Neurosurgical Societies (WFNS) scale and radiomic descriptors of the brain parenchyma as key predictive factors. Additionally, a radiomic model was developed for the automatic estimation of the SEBES (Subarachnoid Hemorrhage Early Brain Edema Score), showing high concordance with expert assessment, albeit with lower prognostic value compared with models specifically optimized for each clinical outcome. Overall, this thesis demonstrates that the integration of ML algorithms with biomarkers derived from HSI and radiomics enables a more objective, reproducible, and efficient quantitative analysis of neurological pathology, although clinical generalization will require prospective multicenter validation. RESUMEN La práctica clínica en neurociencia se enfrenta a limitaciones persistentes en precisión diagnóstica, reproducibilidad y eficiencia. La interpretación visual de la imagen médica es subjetiva, las escalas clínicas ofrecen una caracterización fisiopatológica parcial y la segmentación manual de lesiones es costosa en tiempo y recursos. Estas limitaciones son especialmente críticas en patologías agudas como la hemorragia subaracnoidea aneurismática (HSAa), donde los retrasos en la cuantificación pueden afectar al pronóstico y al manejo terapéutico. En este contexto, esta tesis doctoral desarrolla y valida sistemas automatizados de análisis de imagen médica basados en aprendizaje automático (Machine Learning, ML) para mejorar la objetividad, reproducibilidad y eficiencia del análisis radiológico, así como la predicción de desenlaces clínicos y el apoyo a la toma de decisiones en neurocirugía. El trabajo se articula en tres ejes: (i) clasificación tisular intraoperatoria mediante imagen hiperespectral (HSI) en glioblastoma multiforme (GBM), (ii) segmentación automática de hemorragias intracraneales en tomografía computarizada (TC) en HSAa y (iii) modelización pronóstica basada en biomarcadores radiómicos derivados de segmentaciones automáticas de las hemorragias y del parénquima cerebral adyacente. El primer eje se desarrolla en una cohorte prospectiva de 12 pacientes con GBM intervenidos en el Hospital Universitario 12 de Octubre, con adquisición de 13 cubos HSI mediante un sistema snapshot de 25 bandas (655-975 nm). Se han segmentado cinco clases tisulares y se han entrenado modelos de ML para clasificación tisular. La validación intrapaciente alcanza precisiones globales mayores de 95%, mientras que en el escenario interpaciente (leave-one-patient-out) la precisión global es del 60%, evidenciando la variabilidad biológica. La eficiencia por banda supera la de estudios previos con mayor resolución espectral y los datos se incorporaron a la base...
- Open Access
- 22 March 2026
- Published by Elsevier in Journal of Clinical Neuroscience
- Vol. 149, 111993
Deep learning-based segmentation of aneurysmal subarachnoid hemorrhage: toward accurate and scalable prognostic imaging
- 22 March 2026
- Published by Elsevier in Journal of Clinical Neuroscience
- Vol. 149, 111993
Open AccessDeep learning-based segmentation of aneurysmal subarachnoid hemorrhage: toward accurate and scalable prognostic imaging
- Open Access
Integrating AI-powered automated neurovascular bundle segmentation and radiomics for prostate cancer staging
Background: Assessment of the prostatic neurovascular bundles on MRI is clinically relevant for staging and treatment planning but remains technically challenging and underexplored in automated imaging pipelines. Purpose: To develop and evaluate an automated framework for neurovascular bundle segmentation, proximity-based invasion risk stratification, and radiomics-based prediction of biochemical recurrence, perineural invasion, and extraprostatic extension. Methods: This retrospective study included 808 prostate MRI examinations from three datasets acquired between 2015 and 2020. Among them, 470 T2-weighted image sequences were manually annotated to train a 3D full-resolution nnU-Net segmentation model. Tumor-to-neurovascular bundle distance was used to define invasion risk categories (low, intermediate, high). Machine learning models were developed using radiomics features extracted from neurovascular bundles, lesions, and combined regions, with optional inclusion of prostate-specific antigen and age at MRI. Model performance was evaluated using area under the receiver operating characteristic curve and accuracy. Model interpretability was assessed using Shapley additive explanations. Results: The median patient age was 69 years (interquartile range, 63–73). Automatic neurovascular bundles segmentation achieved anatomically plausible contours, with average surface distance below 1 mm and volume difference under 0.4 cc. The resulting tumor-to-neurovascular bundle invasion risk classification reached 90% accuracy, supporting usability. Radiomics models showed predictive value across endpoints, with moderate testing performance for biochemical recurrence (AUC = 0.73), and higher discrimination for perineural invasion (AUC = 0.80) and extraprostatic extension (AUC = 0.80). Interpretability analysis revealed that tumor-to-neurovascular bundle proximity and neurovascular bundles imaging features among the most relevant contributors to outcome prediction. Conclusions: Automated neurovascular bundle segmentation enabled quantitative tumor proximity assessment and radiomics-based prediction of biochemical recurrence, perineural invasion, and extraprostatic extension.
Open AccessIntegrating AI-powered automated neurovascular bundle segmentation and radiomics for prostate cancer staging
Background: Assessment of the prostatic neurovascular bundles on MRI is clinically relevant for staging and treatment planning but remains technically challenging and underexplored in automated imaging pipelines. Purpose: To develop and evaluate an automated framework for neurovascular bundle segmentation, proximity-based invasion risk stratification, and radiomics-based prediction of biochemical recurrence, perineural invasion, and extraprostatic extension. Methods: This retrospective study included 808 prostate MRI examinations from three datasets acquired between 2015 and 2020. Among them, 470 T2-weighted image sequences were manually annotated to train a 3D full-resolution nnU-Net segmentation model. Tumor-to-neurovascular bundle distance was used to define invasion risk categories (low, intermediate, high). Machine learning models were developed using radiomics features extracted from neurovascular bundles, lesions, and combined regions, with optional inclusion of prostate-specific antigen and age at MRI. Model performance was evaluated using area under the receiver operating characteristic curve and accuracy. Model interpretability was assessed using Shapley additive explanations. Results: The median patient age was 69 years (interquartile range, 63–73). Automatic neurovascular bundles segmentation achieved anatomically plausible contours, with average surface distance below 1 mm and volume difference under 0.4 cc. The resulting tumor-to-neurovascular bundle invasion risk classification reached 90% accuracy, supporting usability. Radiomics models showed predictive value across endpoints, with moderate testing performance for biochemical recurrence (AUC = 0.73), and higher discrimination for perineural invasion (AUC = 0.80) and extraprostatic extension (AUC = 0.80). Interpretability analysis revealed that tumor-to-neurovascular bundle proximity and neurovascular bundles imaging features among the most relevant contributors to outcome prediction. Conclusions: Automated neurovascular bundle segmentation enabled quantitative tumor proximity assessment and radiomics-based prediction of biochemical recurrence, perineural invasion, and extraprostatic extension.
- Open Access
- 15 July 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 270, 108931
In-silico patient-specific modelling of prostate cancer: Predicting PSA dynamics and treatment response
- 15 July 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 270, 108931
Open AccessIn-silico patient-specific modelling of prostate cancer: Predicting PSA dynamics and treatment response
- Open Access
- 24 June 2025
- Published by Springer Nature in Neurosurgical Review
- Vol. 48 (1) , 528
Comprehensive predictive modeling in subarachnoid hemorrhage: integrating radiomics and clinical variables
Subarachnoid hemorrhage (SAH) is a severe condition with high morbidity and long-term neurological consequences. Radiomics, by extracting quantitative features from Computed Tomograhpy (CT) scans, may reveal imaging biomarkers predictive of outcomes. This study evaluates the predictive value of radiomics in SAH for multiple outcomes and compares its performance to models based on clinical data.Radiomic features were extracted from admission CTs using segmentations of brain tissue (white and gray matter) and hemorrhage. Machine learning models with cross-validation were trained using clinical data, radiomics, or both, to predict 6-month mortality, Glasgow Outcome Scale (GOS), vasospasm, and long-term hydrocephalus. SHapley Additive exPlanations (SHAP) analysis was used to interpret feature contributions.The training dataset included 403 aneurysmal SAH patients; GOS predictions used all patients, while vasospasm and hydrocephalus predictions excluded those with incomplete data or early death, leaving 328 and 332 patients, respectively. Radiomics and clinical models demonstrated comparable performance, achieving in validation set AUCs more than 85% for six-month mortality and clinical outcome, and 75% and 86% for vasospasm and hydrocephalus, respectively. In an independent cohort of 41 patients, the combined models yielded AUCs of 89% for mortality, 87% for clinical outcome, 66% for vasospasm, and 72% for hydrocephalus. SHAP analysis highlighted significant contributions of radiomic features from brain tissue and hemorrhage segmentation, alongside key clinical variables, in predicting SAH outcomes.This study underscores the potential of radiomics-based approaches for SAH outcome prediction, demonstrating predictive power comparable to traditional clinical models and enhancing understanding of SAH-related complications. Clinical trial number Not applicable.
- 24 June 2025
- Published by Springer Nature in Neurosurgical Review
- Vol. 48 (1) , 528
Open AccessComprehensive predictive modeling in subarachnoid hemorrhage: integrating radiomics and clinical variables
Subarachnoid hemorrhage (SAH) is a severe condition with high morbidity and long-term neurological consequences. Radiomics, by extracting quantitative features from Computed Tomograhpy (CT) scans, may reveal imaging biomarkers predictive of outcomes. This study evaluates the predictive value of radiomics in SAH for multiple outcomes and compares its performance to models based on clinical data.Radiomic features were extracted from admission CTs using segmentations of brain tissue (white and gray matter) and hemorrhage. Machine learning models with cross-validation were trained using clinical data, radiomics, or both, to predict 6-month mortality, Glasgow Outcome Scale (GOS), vasospasm, and long-term hydrocephalus. SHapley Additive exPlanations (SHAP) analysis was used to interpret feature contributions.The training dataset included 403 aneurysmal SAH patients; GOS predictions used all patients, while vasospasm and hydrocephalus predictions excluded those with incomplete data or early death, leaving 328 and 332 patients, respectively. Radiomics and clinical models demonstrated comparable performance, achieving in validation set AUCs more than 85% for six-month mortality and clinical outcome, and 75% and 86% for vasospasm and hydrocephalus, respectively. In an independent cohort of 41 patients, the combined models yielded AUCs of 89% for mortality, 87% for clinical outcome, 66% for vasospasm, and 72% for hydrocephalus. SHAP analysis highlighted significant contributions of radiomic features from brain tissue and hemorrhage segmentation, alongside key clinical variables, in predicting SAH outcomes.This study underscores the potential of radiomics-based approaches for SAH outcome prediction, demonstrating predictive power comparable to traditional clinical models and enhancing understanding of SAH-related complications. Clinical trial number Not applicable.
- Open Access
- 21 November 2023
- Published by Springer Nature in Research Square
SLIM Brain Database: A Multimodal Image Database of In-vivo Human Brains for Tumour Detection
Hyperspectral imaging (HSI) and machine learning (ML) have been employed in the medical field for classifying highly infiltrative brain tumors. Although existing HSI databases of in-vivo human brains are available, they present two main deficiencies. Firstly, the amount of labeled data is scarce and secondly, 3D-tissue information is unavailable. To address both issues, we present the SLIM Brain database, a multimodal image database of in-vivo human brains which provides HS brain tissue data within the 400-1000 nm spectrum, as well as RGB, depth and multi-view images. Two HS cameras, two depth cameras and different RGB sensors were used to capture images and videos from 193 patients. All data in the SLIM Brain database can be used in a variety of ways, for example to train ML models with more than 1 million HS pixels available and labeled by neurosurgeons, to reconstruct 3D scenes or to visualize RGB brain images with different pathologies, offering unprecedented flexibility for both the medical and engineering communities
- 21 November 2023
- Published by Springer Nature in Research Square
Open AccessSLIM Brain Database: A Multimodal Image Database of In-vivo Human Brains for Tumour Detection
Hyperspectral imaging (HSI) and machine learning (ML) have been employed in the medical field for classifying highly infiltrative brain tumors. Although existing HSI databases of in-vivo human brains are available, they present two main deficiencies. Firstly, the amount of labeled data is scarce and secondly, 3D-tissue information is unavailable. To address both issues, we present the SLIM Brain database, a multimodal image database of in-vivo human brains which provides HS brain tissue data within the 400-1000 nm spectrum, as well as RGB, depth and multi-view images. Two HS cameras, two depth cameras and different RGB sensors were used to capture images and videos from 193 patients. All data in the SLIM Brain database can be used in a variety of ways, for example to train ML models with more than 1 million HS pixels available and labeled by neurosurgeons, to reconstruct 3D scenes or to visualize RGB brain images with different pathologies, offering unprecedented flexibility for both the medical and engineering communities
- 1 January 2023
- Published by Springer Nature
- p. 61-78
How to Extract Radiomic Features from Imaging
- 30 July 2022
- Published by Springer Nature
- p. 123-135
Data-Type Assessment for Real-Time Hyperspectral Classification in Medical Imaging
- 30 July 2022
- Published by Springer Nature
- p. 123-135
Data-Type Assessment for Real-Time Hyperspectral Classification in Medical Imaging
- 24 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 101, 1-6
An Embedded GPU Accelerated Hyperspectral Video Classification System in Real-Time
Uyperspectral imaging (HSI) has been adopted during the last years in different applications where material classification plays an important role. This favoured the improvement and development of new HS sensors, leading to HS snapshot cameras, where the sensor is able to acquire images in a realtime video fashion, reducing the spatial and spectral resolution for these cameras. However, although these cameras are able to acquire RAW HS images at high frames per second (FPS) rates, it is necessary to pre-process them to obtain an actual HS image and then obtain a classification map. This work addresses the development of an embedded CPU+GPU HS video classification system able to acquire and classify HS images in real-time video, i.e., at more than 25 FPS. Also, it includes a comparison where three different CPU+GPU embedded platforms are tested: NVIDIA Jetson Nano, NVIDIA Jetson TX1 and NVIDIA Jetson TX2. The results obtained show the feasibility of a HS video classification system using these embedded platforms.
- 24 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 101, 1-6
An Embedded GPU Accelerated Hyperspectral Video Classification System in Real-Time
Uyperspectral imaging (HSI) has been adopted during the last years in different applications where material classification plays an important role. This favoured the improvement and development of new HS sensors, leading to HS snapshot cameras, where the sensor is able to acquire images in a realtime video fashion, reducing the spatial and spectral resolution for these cameras. However, although these cameras are able to acquire RAW HS images at high frames per second (FPS) rates, it is necessary to pre-process them to obtain an actual HS image and then obtain a classification map. This work addresses the development of an embedded CPU+GPU HS video classification system able to acquire and classify HS images in real-time video, i.e., at more than 25 FPS. Also, it includes a comparison where three different CPU+GPU embedded platforms are tested: NVIDIA Jetson Nano, NVIDIA Jetson TX1 and NVIDIA Jetson TX2. The results obtained show the feasibility of a HS video classification system using these embedded platforms.
- 24 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Stitching technique based on SURF for Hyperspectral Pushbroom Linescan Cameras
Hyperspectral imaging (HSI) has a crucial role in material classification tasks. The primary type of sensors in this field are linescan sensors, offering a high spatial and spectral resolution and improving the quality of the captured hypercubes. However, these cameras have inherent problems such as the need of dedicated hardware to move the camera over the scene and the necessity of pre-processing the raw captures in order to obtain the actual hypercube. This work makes an analysis and an adaptation of stitching techniques to apply them to the medical area. Also, it includes different methods to post-process the hypercubes obtained and it analyzes the SURF algorithm to improve processing times. The results obtained show that a stitching technique based on SURF is suitable for the medical area.
- 24 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Stitching technique based on SURF for Hyperspectral Pushbroom Linescan Cameras
Hyperspectral imaging (HSI) has a crucial role in material classification tasks. The primary type of sensors in this field are linescan sensors, offering a high spatial and spectral resolution and improving the quality of the captured hypercubes. However, these cameras have inherent problems such as the need of dedicated hardware to move the camera over the scene and the necessity of pre-processing the raw captures in order to obtain the actual hypercube. This work makes an analysis and an adaptation of stitching techniques to apply them to the medical area. Also, it includes different methods to post-process the hypercubes obtained and it analyzes the SURF algorithm to improve processing times. The results obtained show that a stitching technique based on SURF is suitable for the medical area.