Effect of upper body venoarterial ECMO on systemic hemodynamics and oxygenation: A computational study.

BACKGROUND: This study seeks to quantify the effects of upper body veno-arterial extracorporeal membrane oxygenation (VA ECMO) on the anatomical distribution of oxygen delivery in the setting of hypoxic respiratory failure and provide new insights that will guide clinical use of this support strategy to bridge patients to lung transplant. METHODS: Employing a patient-specific vascular geometry and a quantitative model of oxygen transport, computational simulations were performed to determine hemodynamics and oxygen delivery in the ascending and descending aorta, left and right coronary arteries, and great vessels during upper body VA ECMO support. Oxygen content in ECMO circuit blood flow was varied while considering different degrees of lung failure severity. Using lumped parameter models to dynamically apply perfusion boundary conditions, hemodynamic parameters and oxygen content were analyzed to assess the effect of ECMO supply titration. RESULTS: The results emphasize the importance of anatomical distribution for tissue oxygen delivery in severe lung failure, with ECMO-derived flow primarily augmenting oxygen content in specific vascular beds. They also demonstrate that although cannulating the subclavian artery can enhance cerebral oxygen delivery, its ability to ensure sufficient oxygen delivery to the coronary circulation seems to be comparatively restricted. CONCLUSIONS: The oxygen delivery to a specific vascular area is primarily determined by the oxygen content in the source of perfusion. Caution is advised with upper body VA ECMO for patients with hypoxic respiratory failure and right ventricle dysfunction, due to potential coronary ischemia. Management of these patients is challenging due to disease progression and organ availability uncertainties.

Modular GAN: positron emission tomography image reconstruction using two generative adversarial networks.

Introduction: The reconstruction of PET images involves converting sinograms, which represent the measured counts of radioactive emissions using detector rings encircling the patient, into meaningful images. However, the quality of PET data acquisition is impacted by physical factors, photon count statistics and detector characteristics, which affect the signal-to-noise ratio, resolution and quantitative accuracy of the resulting images. To address these influences, correction methods have been developed to mitigate each of these issues separately. Recently, generative adversarial networks (GANs) based on machine learning have shown promise in learning the complex mapping between acquired PET data and reconstructed tomographic images. This study aims to investigate the properties of training images that contribute to GAN performance when non-clinical images are used for training. Additionally, we describe a method to correct common PET imaging artefacts without relying on patient-specific anatomical images. Methods: The modular GAN framework includes two GANs. Module 1, resembling Pix2pix architecture, is trained on non-clinical sinogram-image pairs. Training data are optimised by considering image properties defined by metrics. The second module utilises adaptive instance normalisation and style embedding to enhance the quality of images from Module 1. Additional perceptual and patch-based loss functions are employed in training both modules. The performance of the new framework was compared with that of existing methods, (filtered backprojection (FBP) and ordered subset expectation maximisation (OSEM) without and with point spread function (OSEM-PSF)) with respect to correction for attenuation, patient motion and noise in simulated, NEMA phantom and human imaging data. Evaluation metrics included structural similarity (SSIM), peak-signal-to-noise ratio (PSNR), relative root mean squared error (rRMSE) for simulated data, and contrast-to-noise ratio (CNR) for NEMA phantom and human data. Results: For simulated test data, the performance of the proposed framework was both qualitatively and quantitatively superior to that of FBP and OSEM. In the presence of noise, Module 1 generated images with a SSIM of 0.48 and higher. These images exhibited coarse structures that were subsequently refined by Module 2, yielding images with an SSIM higher than 0.71 (at least 22% higher than OSEM). The proposed method was robust against noise and motion. For NEMA phantoms, it achieved higher CNR values than OSEM. For human images, the CNR in brain regions was significantly higher than that of FBP and OSEM (p < 0.05, paired t-test). The CNR of images reconstructed with OSEM-PSF was similar to those reconstructed using the proposed method. Conclusion: The proposed image reconstruction method can produce PET images with artefact correction.

Design of double Z-scheme Ag-Ag3O4/CuO-CuFe2O4 magnetic nanophotocatalyst via starch-templated microwave-combustion hybrid precipitation method and modified with corona-plasma: Remediation of dye contaminants.

Herein, the novel double Z-scheme Ag-Ag3O4/CuO-CuFe2O4 magnetic nanophotocatalyst with nanosphere-on-nanosheet-like morphology was synthesized via the corona-plasma-assisted starch-templated microwave-combustion-precipitation method to remove the dye pollutants. The CuO-CuFe2O4 meso/macroporous nanophotocatalyst was synthesized using a one-pot-stage combustion-microwave process with/without starch as a hard-template. Subsequently, surface modification was carried out by DC corona-plasma discharge technology at various voltages, namely 500, 1000 and 1500 V. Then, the Ag3O4 photocatalyst was deposited on the CuO-CuFe2O4 fabricated with starch-hard-template and treated with 1000 V corona-plasma (denoted as: Ag-Ag3O4/CuO-CuFe2O4 (Starch) 1000 P). The properties of the synthesized nanophotocatalysts were analyzed using various techniques, including X-ray diffraction (XRD), Diffuse reflectance spectroscopy (DRS), Transmission electron microscopy (TEM), Field emission scanning electron microscopy (FESEM), Brunauer-Emmett-Teller and Barrett-Joyner-Halenda (BET-BJH), Vibrating Sample Manetometer (VSM), and Photoluminescence (PL). The XRD analysis corroborated the presence of CuO, CuFe2O4 and Ag3O4 in the structure of all samples. The BET-BJH analysis indicates that the specific surface area of the Ag-Ag3O4/CuO-CuFe2O4 (Starch) 1000 P nanophotocatalyst as the best sample is 2 m2/g, higher than other samples. Additionally, the DRS analysis revealed that the band gap of the Ag-Ag3O4/CuO-CuFe2O4 (Starch) 1000 P nanophotocatalyst is about 1.68 eV with the surface plasmon resonance. The performance of the ternary heterostructured Ag-Ag3O4/CuO-CuFe2O4 (Starch) 1000 P nanophotocatalyst was 96.2% and 89.1% in the degradation of the crystal violet (10 mg/L) and acid orange 7 (10 mg/L), respectively, proving its outstanding degradation capacity.

Automated extraction of the arterial input function from brain images for parametric PET studies.

BACKGROUND: Accurate measurement of the arterial input function (AIF) is crucial for parametric PET studies, but the AIF is commonly derived from invasive arterial blood sampling. It is possible to use an image-derived input function (IDIF) obtained by imaging a large blood pool, but IDIF measurement in PET brain studies performed on standard field of view scanners is challenging due to lack of a large blood pool in the field-of-view. Here we describe a novel automated approach to estimate the AIF from brain images. RESULTS: Total body 18F-FDG PET data from 12 subjects were split into a model adjustment group (n = 6) and a validation group (n = 6). We developed an AIF estimation framework using wavelet-based methods and unsupervised machine learning to distinguish arterial and venous activity curves, compared to the IDIF from the descending aorta. All of the automatically extracted AIFs in the validation group had similar shape to the IDIF derived from the descending aorta IDIF. The average area under the curve error and normalised root mean square error across validation data were - 1.59 ± 2.93% and 0.17 ± 0.07. CONCLUSIONS: Our automated AIF framework accurately estimates the AIF from brain images. It reduces operator-dependence, and could facilitate the clinical adoption of parametric PET.

ParaPET: non-invasive deep learning method for direct parametric brain PET reconstruction using histoimages.

BACKGROUND: The indirect method for generating parametric images in positron emission tomography (PET) involves the acquisition and reconstruction of dynamic images and temporal modelling of tissue activity given a measured arterial input function. This approach is not robust, as noise in each dynamic image leads to a degradation in parameter estimation. Direct methods incorporate into the image reconstruction step both the kinetic and noise models, leading to improved parametric images. These methods require extensive computational time and large computing resources. Machine learning methods have demonstrated significant potential in overcoming these challenges. But they are limited by the requirement of a paired training dataset. A further challenge within the existing framework is the use of state-of-the-art arterial input function estimation via temporal arterial blood sampling, which is an invasive procedure, or an additional magnetic resonance imaging (MRI) scan for selecting a region where arterial blood signal can be measured from the PET image. We propose a novel machine learning approach for reconstructing high-quality parametric brain images from histoimages produced from time-of-flight PET data without requiring invasive arterial sampling, an MRI scan, or paired training data from standard field-of-view scanners. RESULT: The proposed is tested on a simulated phantom and five oncological subjects undergoing an 18F-FDG-PET scan of the brain using Siemens Biograph Vision Quadra. Kinetic parameters set in the brain phantom correlated strongly with the estimated parameters (K1, k2 and k3, Pearson correlation coefficient of 0.91, 0.92 and 0.93) and a mean squared error of less than 0.0004. In addition, our method significantly outperforms (p < 0.05, paired t-test) the conventional nonlinear least squares method in terms of contrast-to-noise ratio. At last, the proposed method was found to be 37% faster than the conventional method. CONCLUSION: We proposed a direct non-invasive DL-based reconstruction method and produced high-quality parametric maps of the brain. The use of histoimages holds promising potential for enhancing the estimation of parametric images, an area that has not been extensively explored thus far. The proposed method can be applied to subject-specific dynamic PET data alone.

A short 18F-FDG imaging window triple injection neuroimaging protocol for parametric mapping in PET.

BACKGROUND: In parametric PET, kinetic parameters are extracted from dynamic PET images. It is not commonly used in clinical practice because of long scan times and the requirement for an arterial input function (AIF). To address these limitations, we designed an 18F-fluorodeoxyglucose (18F-FDG) triple injection dynamic PET protocol for brain imaging with a standard field of view PET scanner using a 24-min imaging window and an input function modeled using measurements from a region of interest placed over the left ventricle. METHODS: To test the protocol in 6 healthy participants, we examined the quality of voxel-based maps of kinetic parameters in the brain generated using the two-tissue compartment model and compared estimated parameter values with previously published values. We also utilized data from a 36-min validation imaging window to compare (1) the modeled AIF against the input function measured in the validation window; and (2) the net influx rate ([Formula: see text]) computed using parameter estimates from the short imaging window against the net influx rate obtained using Patlak analysis in the validation window. RESULTS: Compared to the AIF measured in the validation window, the input function estimated from the short imaging window achieved a mean area under the curve error of 9%. The voxel-wise Pearson's correlation between [Formula: see text] estimates from the short imaging window and the validation imaging window exceeded 0.95. CONCLUSION: The proposed 24-min triple injection protocol enables parametric 18F-FDG neuroimaging with noninvasive estimation of the AIF from cardiac images using a standard field of view PET scanner.

Recent developments in modeling, imaging, and monitoring of cardiovascular diseases using machine learning.

Cardiovascular diseases are the leading cause of mortality, morbidity, and hospitalization around the world. Recent technological advances have facilitated analyzing, visualizing, and monitoring cardiovascular diseases using emerging computational fluid dynamics, blood flow imaging, and wearable sensing technologies. Yet, computational cost, limited spatiotemporal resolution, and obstacles for thorough data analysis have hindered the utility of such techniques to curb cardiovascular diseases. We herein discuss how leveraging machine learning techniques, and in particular deep learning methods, could overcome these limitations and offer promise for translation. We discuss the remarkable capacity of recently developed machine learning techniques to accelerate flow modeling, enhance the resolution while reduce the noise and scanning time of current blood flow imaging techniques, and accurate detection of cardiovascular diseases using a plethora of data collected by wearable sensors.

Aortic Dissection is Determined by Specific Shape and Hemodynamic Interactions.

The aim of this study was to determine whether specific three-dimensional aortic shape features, extracted via statistical shape analysis (SSA), correlate with the development of thoracic ascending aortic dissection (TAAD) risk and associated aortic hemodynamics. Thirty-one patients followed prospectively with ascending thoracic aortic aneurysm (ATAA), who either did (12 patients) or did not (19 patients) develop TAAD, were included in the study, with aortic arch geometries extracted from computed tomographic angiography (CTA) imaging. Arch geometries were analyzed with SSA, and unsupervised and supervised (linked to dissection outcome) shape features were extracted with principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA), respectively. We determined PLS-DA to be effective at separating dissection and no-dissection patients ([Formula: see text]), with decreased tortuosity and more equal ascending and descending aortic diameters associated with higher dissection risk. In contrast, neither PCA nor traditional morphometric parameters (maximum diameter, tortuosity, or arch volume) were effective at separating dissection and no-dissection patients. The arch shapes associated with higher dissection probability were supported with hemodynamic insight. Computational fluid dynamics (CFD) simulations revealed a correlation between the PLS-DA shape features and wall shear stress (WSS), with higher maximum WSS in the ascending aorta associated with increased risk of dissection occurrence. Our work highlights the potential importance of incorporating higher dimensional geometric assessment of aortic arch anatomy in TAAD risk assessment, and in considering the interdependent influences of arch shape and hemodynamics as mechanistic contributors to TAAD occurrence.

Modeling of an Ultrasound System in Targeted Drug Delivery to Abdominal Aortic Aneurysm: A Patient-Specific in Silico Study Based on Ligand-Receptor Binding.

Targeted drug delivery methods have shown a significant impact on enhancing drug delivery efficiency and reducing drug side effects. While various stimuli have been used to promote the drug delivery process, applying ultrasound (US) waves to control drug particles through the human body, noninvasively, has drawn the scientist's attention. However, microcarriers delivery reaches the aneurysmal artery by US waves that exert volumetric forces on blood, and drug carriers, which can therefore affect blood flow patterns and movement pathways of drug carriers, have not yet been studied. In this study, we developed a 3-D patient-specific model of abdominal aortic aneurysm (AAA) to evaluate the effect of US waves in enhancing the drug-containing microbubbles (MBs) adhered on the AAA lumen through ligand-receptor binding. Thus, a focused US (FUS) transducer with a resonance frequency of ~1.1 MHz was added to the geometry. Then, the surface density of MBs (SDM) adhered on the AAA lumen was calculated at peak acoustic pressure of ~1.1, ~2.2, and ~4.3 MPa. Results indicated that increasing the US pressure had a significant impact on improving the MBs adhered to the intended wall, whereby US waves with the maximum pressure of ~4.3 MPa could enhance ~1- [Formula: see text] MBs adhesion ~98% relative to not using the waves. While US waves have the advantage of more SDM adhered to the whole artery wall, they adversely affect the SDM adhered on the critical wall of the abdominal aorta. Furthermore, when the US strength goes up, a reduction occurs in the SDM adhered. This reduction is higher for smaller MBs, which is the mentioned MBs' size and US strength reduced SDM adhesion by about ~50% relative to systemic injection. Therefore, it can be concluded that drug delivery using the US field increases the SDM adhered to the whole AAA wall and decreases the SDM adhered to the critical wall of AAA.

Development of Ultrasensitive Biomimetic Auditory Hair Cells Based on Piezoresistive Hydrogel Nanocomposites.

With an ageing population, hearing disorders are predicted to rise considerably in the following decades. Thus, developing a new class of artificial auditory system has been highlighted as one of the most exciting research topics for biomedical applications. Herein, a design of a biocompatible piezoresistive-based artificial hair cell sensor is presented consisting of a highly flexible and conductive polyvinyl alcohol (PVA) nanocomposite with vertical graphene nanosheets (VGNs). The bilayer hydrogel sensor demonstrates excellent performance to mimic biological hair cells, responding to acoustic stimuli in the audible range between 60 Hz to 20 kHz. The sensor output demonstrates stable mid-frequency regions (∼4-9 kHz), with the greatest sensitivity as high frequencies (∼13-20 kHz). This is somewhat akin to the mammalian auditory system, which has remarkable sensitivity and sharp tuning at high frequencies due to the "active process". This work validates the PVA/VGN sensor as a potential candidate to play a similar functional role to that of the cochlear hair cells, which also operate over a wide frequency domain in a viscous environment. Further characterizations of the sensor show that increasing the sound amplitude results in higher responses from the sensor while taking it to the depth drops the sensor outputs due to attenuation of sound in water. Meanwhile, the acoustic pressure distribution of sound waves is predicted through finite element analysis, whereby the numerical results are in perfect agreement with experimental data. This proof-of-concept work creates a platform for the future design of susceptible, flexible biomimetic sensors to closely mimic the biological cochlea.

A validation study of a virtual-based haptic system for endoscopic sinus surgery training.

BACKGROUND: The development of endoscopic sinus surgery (ESS) training simulators for clinical environment applications has reduced the existing shortcomings in conventional teaching methods, creating a standard environment for trainers and trainees in a more accurate and repeatable fashion. MATERIALS AND METHODS: In this research, the validation study of an ESS training simulator has been addressed. It is important to consider components that guide trainees to improve their hand movements control in the orbital floor removal in an ESS operation. Therefore, we defined three tasks to perform: pre-experiment learning, training, and evaluation. In these tasks, the critical regions introduced in the virtual training environment are forbidden to be touched. Recruiting 20 participants, divided into two groups, we investigated the performance metrics: quality (the percentage of the realism for the generated force for orbital floor removal and the usefulness of the proposed training system for the surgical educational curricula.), efficiency (time, path length), and safety (touching the goal and forbidden wall). RESULTS: All recruited participants answered a post-evaluation questionnaire regarding their perceptions of training system realism, potential educational benefits, and practiced skills. We investigate the differences between groups' performance metrics by utilizing the analysis of variance-Kruskal-Wallis test. Acquired results indicate that training before the actual process of the surgery has a significant effect on the accuracy and validity of the process for surgeons. CONCLUSIONS: Utilizing a standardized environment, trainers and trainees are able to carry out a process with regular features. In addition to traditional education methods, trainees can learn the risk of surgical operations. The training simulators can, also, provide a standard method for assessing the skills of surgical and medical students.

Phenomenological tissue fracture modeling for an Endoscopic Sinus and Skull Base Surgery training system based on experimental data.

The ideal simulator for Endoscopic Sinus and Skull Base Surgery (ESSS) training must be supported by a physical model and provide repetitive behavior in a controlled environment. Development of realistic tissue models is a key part of ESSS virtual reality (VR)-based surgical simulation. Considerable research has been conducted to address haptic or force feedback and propose a phenomenological tissue fracture model for sino-nasal tissue during surgical tool indentation. Mechanical properties of specific sino-nasal regions of the sheep head have been studied in various indentation and relaxation experiments. Tool insertion at different indentation rates into coronal orbital floor (COF) tissue is modeled as a sequence of three events: deformation, fracture, and cutting. The behavior in the deformation phase can be characterized using a non-linear, rate-dependent modified Kelvin-Voigt model. A non-linear model for tissue behavior prior to the fracture point is presented. The overall model shows a non-positive dependency of maximum force on tool indentation rate, which indicates faster tool insertion velocity decreases the maximum final fracture force. The tissue cutting phase has been modeled to characterize the force necessary to slice through the COF. The proposed model in this study can help develop VR-based ESSS base simulators in otolaryngology and ophthalmology surgeries. Such simulators are useful in preoperative planning, accurate surgical simulation, intelligent robotic assistance, and treatment applications.

A Comparison of Doctor/Patient Satisfaction with Aesthetic Outcomes of Rhinoplasty: a Prospective Study.

The tendency toward cosmetic surgeries has been increasing all over the world. These surgeries are conducted to improve the functions related to the involved organ. Moreover, such procedures are supposed to have significant effects on an individual's physical and mental health. This study aims at comparing doctor/patient satisfaction with the aesthetic outcomes of rhinoplasty. In the present retrospective study, 60 patients, including 26 males and 34 females, were investigated. The patients were candidates for a nose job (rhinoplasty). One week before the surgery, the doctor and the patients were provided with questionnaires including information about the nasal hump, nostrils, nose tip size, upward sloping of the nose at the tip, the display of the middle-lower nose, nasal width, the nasal proportion compared with the face, and the general satisfaction. The information on the aforementioned items was collected to record the condition of nose beauty before rhinoplasty. After the surgery, the patients and the doctor were given the same questionnaire following one-month and three-month intervals, so that the aesthetic outcomes resulted from the surgery were evaluated. In comparison to the first month, the surgeons and the patients had a higher level of satisfaction in the third month. The surgeons' satisfaction with the upward sloping of the nose at the tip, as well as the display of the middle-lower nose, was greater compared to the patients. However, the patients' satisfaction with the nostrils was higher than that of the surgeons. Regarding the other factors investigated in the present study, there was no significant difference between the satisfaction of patients and surgeons. For individuals whose educational levels were higher than high school, there was a higher satisfaction level with the nasal hump, the display of the middle-lower nose, and the nasal proportion compared with the face.

Multivariable control of the bifurcation and harmonic perturbations to improve the performance of air-handling units.

In this research, nonlinear dynamics of an air-ehandling unit (AHU) is studied for tracking objectives, in the presence of harmonic perturbations. Three arbitrary realistic set-paths are considered for the indoor temperature and relative humidity. Two controllers based on feedback linearization (FBL) and pole placement approaches are designed to preserve the dynamic system around the desired tracking paths. It is shown that FBL controller works efficiently in bifurcation control and transforms the quasi-periodic limit cycles into the periodic ones (and consequently comfortable indoor conditions). In addition, FBL controller guarantees suppression of larger periodic limit cycles into the smaller ones, while it requires the lower air and cold water flow rates with less oscillatory behavior (in comparison with the pole-placement controller). However, it is observed that FBL controller fails in bifurcation control when the disturbance frequency increases. Re-tuning the dynamic gains of FBL controller is essential under such conditions.

Design and fabrication of a multipurpose thyroid phantom for medical dosimetry and calibration.

A multipurpose anthropomorphic neck phantom was designed and fabricated for use in medical applications. The designed neck phantom is composed of seven elliptic cylindrical slices with a semi-major axis of 14 cm and a semi-minor axis of 12.5 cm, each having the thickness of 2 cm. The thyroid gland, bony part of the neck, and the windpipe were also built inside the neck phantom. For the purpose of medical dosimetry, some holes were drilled inside the phantom to accommodate the thermoluminescence dosemeters with different shapes and dimensions. For testing the quality of images in nuclear medicine, the thyroid gland was built separately to accommodate the radioactive iodine. Finally, the nuclear medicine images were obtained by inserting (131)I in both male and female thyroid parts.

Multivariable robust adaptive sliding mode control of an industrial boiler-turbine in the presence of modeling imprecisions and external disturbances: A comparison with type-I servo controller.

To guarantee the safety and efficient performance of the power plant, a robust controller for the boiler-turbine unit is needed. In this paper, a robust adaptive sliding mode controller (RASMC) is proposed to control a nonlinear multi-input multi-output (MIMO) model of industrial boiler-turbine unit, in the presence of unknown bounded uncertainties and external disturbances. To overcome the coupled nonlinearities and investigate the zero dynamics, input-output linearization is performed, and then the new decoupled inputs are derived. To tackle the uncertainties and external disturbances, appropriate adaption laws are introduced. For constructing the RASMC, suitable sliding surface is considered. To guarantee the sliding motion occurrence, appropriate control laws are constructed. Then the robustness and stability of the proposed RASMC is proved via Lyapunov stability theory. To compare the performance of the purposed RASMC with traditional control schemes, a type-I servo controller is designed. To evaluate the performance of the proposed control schemes, simulation studies on nonlinear MIMO dynamic system in the presence of high frequency bounded uncertainties and external disturbances are conducted and compared. Comparison of the results reveals the superiority of proposed RASMC over the traditional control schemes. RAMSC acts efficiently in disturbance rejection and keeping the system behavior in desirable tracking objectives, without the existence of unstable quasi-periodic solutions.

Adaptive robust control of cancer chemotherapy in the presence of parametric uncertainties: a comparison between three hypotheses.

In this paper, an adaptive robust control strategy is developed for the manipulation of drug usage and consequently the tumor volume in cancer chemotherapy. Three nonlinear mathematical cell-kill models including log-kill hypothesis, Norton-Simon hypothesis and E(max) hypothesis are considered in the presence of uncertainties. The Lyapunov stability theorem is used to investigate the global stability and tracking convergence of the process response. For the first time, performance of the uncertain process is investigated and compared for three nonlinear models. In addition, the effects of treatment period, initial value of tumor volume (carrying capacity) and the uncertainty amount on dynamic system behaviour are studied. Through a comprehensive evaluation, results are presented and compared for three cell-kill models. According to the results, for a wide range of model uncertainties, the adaptive controller guarantees the robust performance. However, for a given treatment period, more variation in drug usage is required as the amount of model uncertainty increases. Moreover, for both the nominal and uncertain models, less drug usage is required as the treatment period increases.

Optimal robust control of drug delivery in cancer chemotherapy: a comparison between three control approaches.

During the drug delivery process in chemotherapy, both of the cancer cells and normal healthy cells may be killed. In this paper, three mathematical cell-kill models including log-kill hypothesis, Norton-Simon hypothesis and Emax hypothesis are considered. Three control approaches including optimal linear regulation, nonlinear optimal control based on variation of extremals and H∞-robust control based on µ-synthesis are developed. An appropriate cost function is defined such that the amount of required drug is minimized while the tumor volume is reduced. For the first time, performance of the system is investigated and compared for three control strategies; applied on three nonlinear models of the process. In additions, their efficiency is compared in the presence of model parametric uncertainties. It is observed that in the presence of model uncertainties, controller designed based on variation of extremals is more efficient than the linear regulation controller. However, H∞-robust control is more efficient in improving robust performance of the uncertain models with faster tumor reduction and minimum drug usage.