Martin Mezl
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Martin Mezl has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
- Radovan Jirik
Institute of Scientific Instruments Czech Academy of Sciences Brno Czechia
10 shared publications - Radim Kolar
Department of Biomedical Engineering, Faculty of Electrical Engineering and Communication Brno University of Technology Brno Czech Republic
7 shared publications Torfinn Taxt
Department of Biomedicine, University of Bergen, Norway
6 shared publicationsKim Nylund
National Centre for Ultrasound in Gastroenterology, Haukeland University Hospital, Bergen, Norway
5 shared publicationsOdd Helge Gilja
National Centre for Ultrasound in Gastroenterology, Haukeland University Hospital, Bergen, Norway; Department of Clinical Medicine, University of Bergen, Norway
5 shared publicationsVratislav Harabis
International Clinical Research Center, Center of Biomedical Engineering, St. Anne’s University Hospital Brno, Czech Republic
4 shared publications
Latest Publications
- Open Access
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 83528-83541
Deep Learning-Enhanced Ultrasound Analysis: Classifying Breast Tumors Using Segmentation and Feature Extraction
Breast cancer remains a significant global health challenge, requiring accurate and effective diagnostic methods for timely treatment. Ultrasound imaging is a valuable diagnostic tool for breast cancer because of its affordability, accessibility, and non-ionizing radiation properties. This study proposes a classification method for breast ultrasound images that integrates segmentation and feature extraction. Initially, ultrasound images are pre-processed to enhance quality and reduce noise, followed by segmentation using the U-Net++ architecture. Feature extraction is then performed using MobileNetV2, and these features are used to train and validate classification models to differentiate between malignant and benign breast masses. The model's performance is assessed using accuracy, precision, recall, Mean IoU, and Dice Score metrics. The U-Net++ model achieved superior segmentation performance with a Dice Score of 0.911 and a Mean IoU of 0.838, outperforming related methods such as U-Net (0.888 Dice, 0.79 IoU) and Efficient U-Net (0.904 Dice, 0.80 IoU). In the classification task, MobileNetV2 when paired with the ANN classifier, produced the highest test accuracy at 96.58%, with a precision of 97% and recall of 96%. Our approach demonstrates superior performance compared to other models, such as RMTL-Net, which achieved 91.02% accuracy, and hybrid CAD models with 94% accuracy. This highlights the benefits of combining advanced segmentation and feature extraction techniques, with MobileNetV2 proving to be the better model, offering superior accuracy and robustness in classification tasks. This approach has the potential to support promise for supporting radiologists, enhance diagnostic accuracy, and ultimately improve outcomes for breast cancer patients. In the future, we will use comprehensive datasets to validate our methodology.
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 83528-83541
Open AccessDeep Learning-Enhanced Ultrasound Analysis: Classifying Breast Tumors Using Segmentation and Feature Extraction
Breast cancer remains a significant global health challenge, requiring accurate and effective diagnostic methods for timely treatment. Ultrasound imaging is a valuable diagnostic tool for breast cancer because of its affordability, accessibility, and non-ionizing radiation properties. This study proposes a classification method for breast ultrasound images that integrates segmentation and feature extraction. Initially, ultrasound images are pre-processed to enhance quality and reduce noise, followed by segmentation using the U-Net++ architecture. Feature extraction is then performed using MobileNetV2, and these features are used to train and validate classification models to differentiate between malignant and benign breast masses. The model's performance is assessed using accuracy, precision, recall, Mean IoU, and Dice Score metrics. The U-Net++ model achieved superior segmentation performance with a Dice Score of 0.911 and a Mean IoU of 0.838, outperforming related methods such as U-Net (0.888 Dice, 0.79 IoU) and Efficient U-Net (0.904 Dice, 0.80 IoU). In the classification task, MobileNetV2 when paired with the ANN classifier, produced the highest test accuracy at 96.58%, with a precision of 97% and recall of 96%. Our approach demonstrates superior performance compared to other models, such as RMTL-Net, which achieved 91.02% accuracy, and hybrid CAD models with 94% accuracy. This highlights the benefits of combining advanced segmentation and feature extraction techniques, with MobileNetV2 proving to be the better model, offering superior accuracy and robustness in classification tasks. This approach has the potential to support promise for supporting radiologists, enhance diagnostic accuracy, and ultimately improve outcomes for breast cancer patients. In the future, we will use comprehensive datasets to validate our methodology.
- Top Cited Papers
- 10 March 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 22 (6) , 3214-3233
Multimodal Features for Detection of Driver Stress and Fatigue: Review
Driver fatigue and stress significantly contribute to higher number of car accidents worldwide. Although, different detection approaches have been already commercialized and used by car producers (and third party companies), research activities in this field are still needed in order to increase the reliability of these alert systems. Also, in the context of automated driving, the driver mental state assessment will be an important part of cars in future. This paper presents state-of-the-art review of different approaches for driver fatigue and stress detection and evaluation. We describe in details various signals (biological, car and video) and derived features used for these tasks and we discuss their relevance and advantages. In order to make this review complete, we also describe different datasets, acquisition systems and experiment scenarios.
- 10 March 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 22 (6) , 3214-3233
Top Cited PapersMultimodal Features for Detection of Driver Stress and Fatigue: Review
Driver fatigue and stress significantly contribute to higher number of car accidents worldwide. Although, different detection approaches have been already commercialized and used by car producers (and third party companies), research activities in this field are still needed in order to increase the reliability of these alert systems. Also, in the context of automated driving, the driver mental state assessment will be an important part of cars in future. This paper presents state-of-the-art review of different approaches for driver fatigue and stress detection and evaluation. We describe in details various signals (biological, car and video) and derived features used for these tasks and we discuss their relevance and advantages. In order to make this review complete, we also describe different datasets, acquisition systems and experiment scenarios.
- Open Access
- 15 May 2018
- Published by Springer Nature in BMC Medical Imaging
- Vol. 18 (1) , 1-8
Contrast-enhanced ultrasonography of the pancreas shows impaired perfusion in pancreas insufficient cystic fibrosis patients
Perfusion assessment of the pancreas is challenging and poorly evaluated. Pancreatic affection is a prevalent feature of cystic fibrosis (CF). Little is known about pancreatic perfusion in CF. We aimed to assess pancreatic perfusion by contrast-enhanced ultrasound (CEUS) analysed in the bolus-and-burst model and software. We performed contrast enhanced ultrasound of the pancreas in 25 CF patients and 20 healthy controls. Perfusion data was analysed using a dedicated perfusion model providing the mean capillary transit-time (MTT), blood flow (BF) and blood-volume (BV). CF patients were divided according to exocrine function. The pancreas insufficient CF patients had longer MTT (p ≤ 0.002), lower BF (p < 0.001) and lower BV (p < 0.05) compared to the healthy controls and sufficient CF patients. Interrater analysis showed substantial agreement for the analysis of mean transit time. The bolus-and-burst method used on pancreatic CEUS-examinations demonstrates reduced perfusion in CF patients with pancreas affection. The perfusion model and software requires further optimization and standardization to be clinical applicable for the assessment of pancreatic perfusion.
- 15 May 2018
- Published by Springer Nature in BMC Medical Imaging
- Vol. 18 (1) , 1-8
Open AccessContrast-enhanced ultrasonography of the pancreas shows impaired perfusion in pancreas insufficient cystic fibrosis patients
Perfusion assessment of the pancreas is challenging and poorly evaluated. Pancreatic affection is a prevalent feature of cystic fibrosis (CF). Little is known about pancreatic perfusion in CF. We aimed to assess pancreatic perfusion by contrast-enhanced ultrasound (CEUS) analysed in the bolus-and-burst model and software. We performed contrast enhanced ultrasound of the pancreas in 25 CF patients and 20 healthy controls. Perfusion data was analysed using a dedicated perfusion model providing the mean capillary transit-time (MTT), blood flow (BF) and blood-volume (BV). CF patients were divided according to exocrine function. The pancreas insufficient CF patients had longer MTT (p ≤ 0.002), lower BF (p < 0.001) and lower BV (p < 0.05) compared to the healthy controls and sufficient CF patients. Interrater analysis showed substantial agreement for the analysis of mean transit time. The bolus-and-burst method used on pancreatic CEUS-examinations demonstrates reduced perfusion in CF patients with pancreas affection. The perfusion model and software requires further optimization and standardization to be clinical applicable for the assessment of pancreatic perfusion.
- 1 October 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Evaluation of accuracy of bolus and burst method for quantitative ultrasound perfusion analysis with various arterial input function models
Quantitative ultrasound perfusion imaging is used to estimate tissue perfusion parameters, such as blood flow, blood volume and mean transit time. The bolus & burst method is one possible approach to quantitative ultrasound perfusion imaging. It has been presented on clinical and preclinical data with two models of the arterial input function. In this study, the two models are compared on simulated and preclinical data to assess the achievable precision and reproducibility.
- 1 October 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Evaluation of accuracy of bolus and burst method for quantitative ultrasound perfusion analysis with various arterial input function models
Quantitative ultrasound perfusion imaging is used to estimate tissue perfusion parameters, such as blood flow, blood volume and mean transit time. The bolus & burst method is one possible approach to quantitative ultrasound perfusion imaging. It has been presented on clinical and preclinical data with two models of the arterial input function. In this study, the two models are compared on simulated and preclinical data to assess the achievable precision and reproducibility.
- 1 August 2015
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 63, 229-237
Semi-automatic motion compensation of contrast-enhanced ultrasound images from abdominal organs for perfusion analysis
- 1 August 2015
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 63, 229-237
Semi-automatic motion compensation of contrast-enhanced ultrasound images from abdominal organs for perfusion analysis
- 11 May 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
- Vol. 62 (5) , 983-987
Absolute ultrasound perfusion parameter quantification of a tissue-mimicking phantom using bolus tracking [Correspondence]
This study presents three methods for absolute quantification in ultrasound perfusion analysis based on bolus tracking. The first two methods deconvolve the perfusion time sequence with a measured AIF, using a nonparametric or a parametric model of the tissue residue function, respectively. The third method is a simplified approach avoiding deconvolution by assuming a narrow AIF. A phantom with a dialyzer filter as a tissue-mimicking model was used for evaluation. Estimated mean transit times and blood volumes were compared with the theoretical values. A match with a maximum error of 12% was achieved.
- 11 May 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
- Vol. 62 (5) , 983-987
Absolute ultrasound perfusion parameter quantification of a tissue-mimicking phantom using bolus tracking [Correspondence]
This study presents three methods for absolute quantification in ultrasound perfusion analysis based on bolus tracking. The first two methods deconvolve the perfusion time sequence with a measured AIF, using a nonparametric or a parametric model of the tissue residue function, respectively. The third method is a simplified approach avoiding deconvolution by assuming a narrow AIF. A phantom with a dialyzer filter as a tissue-mimicking model was used for evaluation. Estimated mean transit times and blood volumes were compared with the theoretical values. A match with a maximum error of 12% was achieved.
- 1 August 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
- Vol. 2014 (1094687X) , 4276-4279
Blind deconvolution in dynamic contrast-enhanced MRI and ultrasound
This paper is focused on quantitative perfusion analysis using MRI and ultrasound. In both MRI and ultrasound, most approaches allow estimation of rate constants (Ktrans, kep for MRI) and indices (AUC, TTP) that are only related to the physiological perfusion parameters of a tissue (e.g. blood flow, vessel permeability) but do not allow their absolute quantification. Recent methods for quantification of these physiological perfusion parameters are shortly reviewed. The main problem of these methods is estimation of the arterial input function (AIF). This paper summarizes and extends the current blind-deconvolution approaches to AIF estimation. The feasibility of these methods is shown on a small preclinical study using both MRI and ultrasound.
- 1 August 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
- Vol. 2014 (1094687X) , 4276-4279
Blind deconvolution in dynamic contrast-enhanced MRI and ultrasound
This paper is focused on quantitative perfusion analysis using MRI and ultrasound. In both MRI and ultrasound, most approaches allow estimation of rate constants (Ktrans, kep for MRI) and indices (AUC, TTP) that are only related to the physiological perfusion parameters of a tissue (e.g. blood flow, vessel permeability) but do not allow their absolute quantification. Recent methods for quantification of these physiological perfusion parameters are shortly reviewed. The main problem of these methods is estimation of the arterial input function (AIF). This paper summarizes and extends the current blind-deconvolution approaches to AIF estimation. The feasibility of these methods is shown on a small preclinical study using both MRI and ultrasound.
- 1 July 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 541-544
Quantitative myocardial perfusion analysis with contrast-enhanced ultrasound bolus tracking - preliminary animal results
This study aims at absolute quantification in ultra-sound perfusion analysis in vivo. The main task of the study is to apply a perfusion model for intravascular tracers known mainly from other modalities to ultrasound perfusion analysis. Parametric formulation of the tissue residual function is proposed. Three perfusion parameters - blood flow, blood volume and mean transit time are estimated. The method is tested on preclinical data acquired from pig's heart and compared with reference blood flow values obtained using fluorescent microspheres. A fairly good match between ultrasound perfusion and the reference method was achieved.
- 1 July 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 541-544
Quantitative myocardial perfusion analysis with contrast-enhanced ultrasound bolus tracking - preliminary animal results
This study aims at absolute quantification in ultra-sound perfusion analysis in vivo. The main task of the study is to apply a perfusion model for intravascular tracers known mainly from other modalities to ultrasound perfusion analysis. Parametric formulation of the tissue residual function is proposed. Three perfusion parameters - blood flow, blood volume and mean transit time are estimated. The method is tested on preclinical data acquired from pig's heart and compared with reference blood flow values obtained using fluorescent microspheres. A fairly good match between ultrasound perfusion and the reference method was achieved.
- 1 July 2013
- Published by Elsevier in Ultrasound in Medicine & Biology
- Vol. 39 (7) , 1197-1206
Quantitative Contrast-Enhanced Ultrasound Comparison Between Inflammatory and Fibrotic Lesions in Patients with Crohn's Disease
- 1 July 2013
- Published by Elsevier in Ultrasound in Medicine & Biology
- Vol. 39 (7) , 1197-1206
Quantitative Contrast-Enhanced Ultrasound Comparison Between Inflammatory and Fibrotic Lesions in Patients with Crohn's Disease
- 25 January 2013
- Published by IOP Publishing in Physiological Measurement
- Vol. 34 (2) , 151-162
Comparison and evaluation of indicator dilution models for bolus of ultrasound contrast agents
Dynamic contrast-enhanced ultrasound (DCE-US) imaging is a promising diagnostic method, which enables the evaluation of tissue perfusion via different parameters. The mean transit time and time-to-peak parameters are the main time parameters and their values depend on the model used for the approximation of the noisy perfusion curves. In this paper, we described a new comparison of different perfusion models using a tissue mimicking phantom. The following models were compared: log-normal, lagged, Erlang, Gamma and the local density random walk model. We discovered that the mean-square error is not the best criterion for model evaluation. More important is the comparison between the estimated time perfusion parameters and the physical parameters of the developed tissue mimicking phantom. Based on the statistical analysis, we can suggest that for the DCE-US perfusion analysis more models should be used, excluding the log-normal model, which gives the highest error of mean transit time value.
- 25 January 2013
- Published by IOP Publishing in Physiological Measurement
- Vol. 34 (2) , 151-162
Comparison and evaluation of indicator dilution models for bolus of ultrasound contrast agents
Dynamic contrast-enhanced ultrasound (DCE-US) imaging is a promising diagnostic method, which enables the evaluation of tissue perfusion via different parameters. The mean transit time and time-to-peak parameters are the main time parameters and their values depend on the model used for the approximation of the noisy perfusion curves. In this paper, we described a new comparison of different perfusion models using a tissue mimicking phantom. The following models were compared: log-normal, lagged, Erlang, Gamma and the local density random walk model. We discovered that the mean-square error is not the best criterion for model evaluation. More important is the comparison between the estimated time perfusion parameters and the physical parameters of the developed tissue mimicking phantom. Based on the statistical analysis, we can suggest that for the DCE-US perfusion analysis more models should be used, excluding the log-normal model, which gives the highest error of mean transit time value.