Table of contents

This review focuses on recently developed printable biomaterials for bone and mineralized tissue engineering. 3D printing or bioprinting is an advanced technology to design and fabricate complex functional 3D scaffolds, mimicking native tissue for in vivo applications. We categorized the biomaterials into two main classes: 3D printing and bioprinting. Various biomaterials, including natural, synthetic biopolymers and their composites, have been studied. Biomaterial inks or bioinks used for bone and mineralized tissue regeneration include hydrogels loaded with minerals or bioceramics, cells, and growth factors. In 3D printing, the scaffold is created by acellular biomaterials (biomaterial inks), while in 3D bioprinting, cell-laden hydrogels (bioinks) are used. Two main classes of bioceramics, including bioactive and bioinert ceramics, are reviewed. Bioceramics incorporation provides osteoconductive properties and induces bone formation. Each biopolymer and mineral have its advantages and limitations. Each component of these composite biomaterials provides specific properties, and their combination can ameliorate the mechanical properties, bioactivity, or biological integration of the 3D printed scaffold. Present challenges and future approaches to address them are also discussed.

Local post-stenotic hemodynamics has critical influence in the atherosclerotic plaque progression occurring in susceptible arterial sites, in particular the left main coronary artery (LMCA) bifurcation. Understanding the effects of plaque morphological characteristics: stenosis severity (SS), eccentricity index (EI) and lesion length (LL) on the post-stenotic flow behavior can significantly improve treatment planning. In order to investigate these effects, we have employed computational fluid dynamics (CFD) simulations in twenty computer-generated and five patient-specific LMCA models and the hemodynamic parameters: velocity, pressure (P), wall pressure gradient (WPG), wall shear stress (WSS), time averaged wall shear stress (TAWSS), oscillatory shear index (OSI), relative residence time (RRT) and helicity intensity (h₂) were analyzed. Our results revealed that the effect of stenosis eccentricity varied significantly for different values of stenosis severity and lesion length. Regions with low WSS, low TAWSS and high RRT were more prominent in models having higher stenosis severity. For smaller lesion length, at low and moderate stenosis severity, surface area with low TAWSS and high RRT decreased with increasing eccentricity index, whereas for high stenosis severity models, low TAWSS region and average RRT values increased with eccentricity. However, for models with longer lesion length, regions with high OSI and RRT overall increased gradually with eccentricity. The helicity intensity (h₂) of all models remained very low except at the most eccentric model with longer lesion length. The presence of very high helical flow in this model suggests the possibility of atheroprotective flow. It can be concluded that all plaque morphological characteristics covered under this investigation play an important role in plaque progression.

Objectives. anatomical changes are inevitable during the course of radiotherapy treatments and, if significant, can severely alter expected dose distributions and affect treatment outcome. Adaptive radiotherapy (ART) is employed to maintain the planned distribution and minimise detriment to predicted treatment outcome. Typically, patients who may benefit from adaptive planning are identified via a re-planning process, i.e., re-simulation, re-contouring, re-planning and treatment plan quality assurance (QA). This time-intensive process significantly increases workload, can introduce delays and increases unnecessary stress to those patients who will not actually gain benefit. We consider it crucial to develop efficient models to predict changes to target coverage and trigger ART, without the need for re-planning. Methods. knowledge-based planning (KBP) models were developed using data for 20 patients' (400 fractions) to predict changes in PTV V₉₅ coverage $\left({\rm{\Delta }}V{95}^{PTV}\right).$ Initially, this change in coverage was calculated on the synthetic computerised tomography (sCT) images produced using the Velocity adaptive radiotherapy software. Models were developed using patient (cell death bio-marker) and treatment fraction (PTV characteristic) specific parameters to predict $\left({\rm{\Delta }}V{95}^{PTV}\right)$and verified using five patients (100 fractions) data. Results. three models were developed using combinations of patient and fraction specific terms. The prediction accuracy of the model developed using biomarker (PD-L1 expression) and the difference in 'planning' and 'fraction' PTV centre of the mass (characterised by mean square difference, MSD) had the higher prediction accuracy, predicting the $\left({\rm{\Delta }}V{95}^{PTV}\right)$within ± 1.0% for 77% of the total fractions; with 59% for the model developed using, PTV size, PD-L1 and MSD and 48% PTV size and MSD respectively. Conclusion. the KBP models can predict $\left({\rm{\Delta }}V{95}^{PTV}\right)$very effectively and efficiently for advanced-stage NSCLC patients treated using volumetric modulated arc therapy and to identify patients who may benefit from adaption for a specific fraction.

Introduction. Event-related desynchronization (ERD) is used in brain-computer interfaces (BCI) to detect the user's motor intention (MI) and convert it into a command for an actuator to provide sensory feedback or mobility, for example by means of functional electrical stimulation (FES). Recent studies have proposed to evoke the nociceptive withdrawal reflex (NWR) using FES, in order to evoke synergistic movements of the lower limb and to facilitate the gait rehabilitation of stroke patients. The use of NWR to provide sensorimotor feedback in ERD-based BCI is novel; thererfore, the conditioning effect that nociceptive stimuli might have on MI is still unknown. Objetive. To assess the ERD produced during the MI after FES-evoked NWR, in order to evaluate if nociceptive stimuli condition subsequent ERDs. Methods. Data from 528 electroencephalography trials of 8 healthy volunteers were recorded and analyzed. Volunteers used an ERD-based BCI, which provided two types of feedback: intrisic by the FES-evoked NWR and extrinsic by virtual reality. The electromyogram of the tibialis anterior muscle was also recorded. The main outcome variables were the normalized root mean square of the evoked electromyogram (RMS_norm), the average electroencephalogram amplitude at the ERD frequency during MI (${\bar{A}}_{MI}$) and the percentage decrease of ${\bar{A}}_{MI}$ relative to rest ($ERD \%$) at the first MI subsequent to the activation of the BCI. Results. No evidence of changes of the RMS_norm on both the ${\bar{A}}_{MI}$ (p = 0.663) and the $ERD \%$ (p = 0.252) of the subsequent MI was detected. A main effect of the type of feedback was found in the subsequent ${\bar{A}}_{MI}$ (p < 0.001), with intrinsic feedback resulting in a larger ${\bar{A}}_{MI}.$Conclusions. No evidence of ERD conditioning was observed using BCI feedback based on FES-evoked NWR . Significance. FES-evoked NWR could constitute a potential feedback modality in an ERD-based BCI to facilitate motor recovery of stroke people.

Purpose. Metastatic complications are responsible for 90% of cancer-associated mortality. Magnetic resonance imaging (MRI) can be used to observe the brain's microstructure and potentially correlate changes with metastasis occurrence. Diffusion weighted imaging (DWI) is an MRI technique that utilizes the kinetics of water molecules within the body. The aim of this study is to use DWI to characterize diffusion changes within brain metastases in cancer patients pre- and post-stereotactic radiosurgery (SRS). Methods. We retrospectively analyzed 113 metastases from 13 patients who underwent SRS for brain metastasis recurrence. Longitudinal apparent diffusion coefficient (ADC) maps were registered to Gd-T1 images and CT, and clinical metastasis ROIs from all SRS treatments were retrospectively transferred onto these ADC maps for analysis. Metastases were characterized based on pre-SRS diffusion pattern, primary cancer site, and post-SRS outcome. ADC values were calculated pre- and post-SRS. Results.ADC values were significantly elevated (980.2 × 10⁻⁶ mm² s⁻¹ and 1040.3 × 10⁻⁶ mm² s⁻¹ pre- and post-SRS, respectively) when compared to healthy brain tissue (826.8 × 10⁻⁶ mm² s⁻¹) for all metastases. Three identified pre-SRS patterns were significantly different before SRS and within 6 months post-SRS. No significant differences were observed between different primaries pre-SRS. Post-SRS, Lung metastases ADC decreased by 86.2 × 10⁻⁶ mm² s⁻¹, breast metastases increased by 116.7 × 10⁻⁶ mm² s⁻¹, and genitourinary metastases showed no significant ADC change. SRS outcomes showed ADC variability pre-treatment but no significant differences pre- and post-SRS, except at 6–9 months post-SRS where progressing metastases were elevated when compared to other response groups. Conclusion. This study provided a unique opportunity to characterize diffusion changes in brain metastases before their manifestation on standard Gd-T1 images and post-SRS. Identified patterns may improve early detection of brain metastases as well as predict their response to treatment.

At the diagnostic reference level (DRL) related to medical radiation, DRL quantity for general radiography is the entrance surface dose (ESD). Calculation of the ESD in medical radiography requires the backscatter factor (BSF), but derivation of the BSF requires assessment of an irradiated simulation of a human body. The present study used optically stimulated luminescence (OSL) dosimeters and an anthropomorphic phantom as the irradiated body, and the BSF was calculated for different half value layer (HVL)s and field sizes. The need for different BSFs for different regions was also investigated by derivationing of the BSFs for different regions. The pelvis of a RANDO phantom was irradiated under the conditions of the HVL of 2.0, 3.1, and 4.6 mmAl; tube current of 200 mA; irradiation time of 0.1 s; source surface distance of 100 cm; and field sizes of 10 × 10 cm², 20 cm², 30 cm², and 40 cm². Measurement in air was performed under the same conditions. Several threads were stretched through the air with tissue paper placed on them and the nanoDot dosimeters placed on the paper. Four dosimeters were placed, and measurement was performed 5 times under each set of conditions. The compared radiographed regions were the skull, chest, and pelvis. The BSF increased with increasing HVL size and with increasing field size. The larger the HVL, the larger the difference between field sizes of 10 × 10 cm² and 40 × 40 cm² and the larger the increase in BSF relative to the increase in field size. The BSF differed by region, from large to small in the order chest, pelvis, and skull. The results thus showed that the BSF differs by the radiographed region. Thus, it is desirable to determine the BSF in each radiographed region by investigation with an anthropomorphic phantom.

Attenuation correction of annihilation photons is essential in PET image reconstruction for providing accurate quantitative activity maps. In the absence of an aligned CT device to obtain attenuation information, we propose the high-resolution residual U-net (HRU-Net) to extract attenuation correction factors (ACF) directly from time-of-flight (TOF) PET emission data. HRU-Net is built upon the U-Net encoding-decoding architecture and it utilizes four blocks of modified residual connections in each stage. In each residual block, concatenation is performed to incorporate input and output feature vectors. In addition, flexible and efficient elements of convolutional neural network (CNN) such as dilated convolutions, pre-activation order of a batch normalization (BN) layer, a rectified linear unit (ReLU) layer and a convolution layer, and residual connections are utilized to extract high resolution features. To illustrate the effectiveness of the proposed method, HRU-Net estimated ACF, attenuation maps and activity maps are compared with maximum likelihood ACF (MLACF) algorithm, U-Net, and HC-Net. An ablation study is conducted using non-TOF and TOF sinograms as inputs of networks. The experimental results show that HRU-Net with TOF projections as inputs leads to normalized root mean square error (NRMSE) of 4.84% $\pm$ 1.58%, outperforming MLACF, U-Net and HC-Net with NRMSE of 47.82% $\pm$ 13.62%, 6.92% $\pm$ 1.94%, and 7.99% $\pm$ 2.49%, respectively.

This paper presents a novel two-dimensional (2D) computational modeling of the retinal oxygen delivery, transport, and consumption analysis. The 2D modeling allows the division of the retina into four layers to address different flow profiles. The retina domain was meshed using the ICEM CFD mesher, while the ANSYS Fluent was used to calculate the transport phenomena for the four different layers. Clinical cases such as diabetes and sickle cell anaemia denoting the effects of decreasing retinal blood flow and hemoglobin's oxygen affinity were investigated. The simulation results showed that for a healthy retina in light and dark conditions, the outer retina is in danger of hypoxia at thickness >197.56 μm. However, the treatment of severe ischaemia using extreme hyperoxia seems beneficial for retinal thickness >197.56 μm but harmful for thickness <122.75 μm. The reduction of hemoglobin's oxygen affinity at low blood flow regimes could not improve the retina's oxygen levels. The study supports the oxygen toxicity hypothesis that hypoxia causes retina degeneration and estimates the retinal thickness and lighting conditions (dark or light) this may occur.

Positron Emission Tomography (PET) reconstructed image signal-to-noise ratio (SNR) can be improved by including the 511 keV photon pair coincidence time-of-flight (TOF) information. The degree of SNR improvement from this TOF capability depends on the coincidence time resolution (CTR) of the PET system, which is essentially the variation in photon arrival time differences over all coincident photon pairs detected for a point positron source placed at the system center. The CTR is determined by several factors including the intrinsic properties of the scintillation crystals and photodetectors, crystal-to-photodetector coupling configurations, reflective materials, and the electronic readout configuration scheme. The goal of the present work is to build a novel TOF-PET system with 100 picoseconds (ps) CTR, which provides an additional factor of 1.5–2.0 improvement in reconstructed image SNR compared to state-of-the-art TOF-PET systems which achieve 225–400 ps CTR. A critical parameter to understand is the optical reflector's influence on scintillation light collection and transit time variations to the photodetector. To study the effects of the reflector covering the scintillation crystal element on CTR, we have tested the performance of four different reflector materials: Enhanced Specular Reflector (ESR) –coupled with air or optical grease to the scintillator; Teflon tape; BaSO₄ paint alone or mixed with epoxy; and TiO₂ paint. For the experimental set-up, we made use of 3 × 3 × 10 mm³ fast-LGSO:Ce scintillation crystal elements coupled to an array of silicon photomultipliers (SiPMs) using a novel 'side-readout' configuration that has proven to have lower variations in scintillation light collection efficiency and transit time to the photodetector. Results: show CTR values of 102.0 ± 0.8, 100.2 ± 1.2, 97.3 ± 1.8 and 95.0 ± 1.0 ps full-width-half-maximum (FWHM) with non-calibrated energy resolutions of 10.2 ± 1.8, 9.9 ± 1.2, 7.9 ± 1.2, and 8.6 ± 1.7% FWHM for the Teflon, ESR (without grease), BaSO₄ (without epoxy) and TiO₂ paint treatments, respectively.

High frequency electrical stimulation of brain is commonly used in research experiments and clinical trials as a modern tool for control of epileptic seizures. However, the mechanistic basis by which periodic external stimuli alter the brain state is not well understood. This study provides a computational insight into the mechanism of seizure suppression by high frequency stimulation (HFS). In particular, a modified version of the Jansen-Rit neural mass model is employed, in which EEG signals can be considered as the input. The proposed model reproduces seizure-like activity in the output during the ictal period of the input signal. By applying a control signal to the model, a wide range of stimulation amplitudes and frequencies are systematically explored. Simulation results reveal that HFS can effectively suppress the seizure-like activity. Our results suggest that HFS has the ability of shifting the operating state of neural populations away from a critical condition. Furthermore, a closed-loop control strategy is proposed in this paper. The main objective has been to considerably reduce the control effort needed for blocking abnormal activity of the brain. Such an energy reduction could be of practical importance, to reduce possible side effects and increase battery life for implanted neurostimulators.

In previous works, we showed that incorporating individual airways as organs-at-risk (OARs) in the treatment of lung stereotactic ablative radiotherapy (SAbR) patients potentially mitigates post-SAbR radiation injury. However, the performance of common clinical dose calculation algorithms in airways has not been thoroughly studied. Airways are of particular concern because their small size and the density differences they create have the potential to hinder dose calculation accuracy. To address this gap in knowledge, here we investigate dosimetric accuracy in airways of two commonly used dose calculation algorithms, the anisotropic analytical algorithm (AAA) and Acuros-XB (AXB), recreating clinical treatment plans on a cohort of four SAbR patients. A virtual bronchoscopy software was used to delineate 856 airways on a high-resolution breath-hold CT (BHCT) image acquired for each patient. The planning target volumes (PTVs) and standard thoracic OARs were contoured on an average CT (AVG) image over the breathing cycle. Conformal and intensity-modulated radiation therapy plans were recreated on the BHCT image and on the AVG image, for a total of four plan types per patient. Dose calculations were performed using AAA and AXB, and the differences in maximum and mean dose in each structure were calculated. The median differences in maximum dose among all airways were ≤0.3Gy in magnitude for all four plan types. With airways grouped by dose-to-structure or diameter, median dose differences were still ≤0.5Gy in magnitude, with no clear dependence on airway size. These results, along with our previous airway radiosensitivity works, suggest that dose differences between AAA and AXB correspond to an airway collapse variation ≤0.7% in magnitude. This variation in airway injury risk can be considered as not clinically relevant, and the use of either AAA or AXB is therefore appropriate when including patient airways as individual OARs so as to reduce risk of radiation-induced lung toxicity.

In this paper, we study the reversible electroporation process on normal and cancerous cervical cells. The 2D contour of the cervical cells is extracted using image processing techniques from the Pap smear images. The conductivity change in the cancer cell model has been used to differentiate the effects of the high-frequency electric field on normal and cancerous cells. The cells' dielectric constant modulates when this high-frequency pulse is applied based on the Debye relaxation. To computationally visualize the effects of the electroporation on the cell membrane, the Smoluchowski equation is employed to estimate pore density, and Maxwell equations are used to determine the electric potential developed across the membrane of the cervical cell. The results demonstrate the suitability of this mathematical model for studying the response of normal and cancerous cells under electric stress. The electric field is supplied with the help of a realistic pulse generator which is designed on the principle of Marx circuit and avalanche transistor-based operations to produce a Gaussian pulse. The paper here uses a strength-duration curve to differentiate the electric field and time in nanoseconds required to electroporate normal and cancerous cells.

Grasping of the objects is the most frequent activity performed by the human upper limb. The amputations of the upper limb results in the need for prosthetic devices. The myoelectric prosthetic devices use muscle signals and apply control techniques for identification of different levels of hand gesture and force levels. In this study; a different level force contraction experiment was performed in which Electromyography (EMGs) signals and fingertip force signals were acquired. Using this experimental data; a two-step feature selection process is applied for the designing of a pattern recognition algorithm for the classification of different force levels. The two step feature selection process consist of generalized feature ranking using ReliefF, followed by personalized feature selection using Neighborhood Component Analysis (NCA) from the shortlisted features by earlier technique. The classification algorithms applied in this study were Support Vector Machines (SVM) and Random Forest (RF). Besides feature selection; optimization of the number of muscles during classification of force levels was also performed using designed algorithm. Based on this algorithm; the maximum classification accuracy using SVM classifier and two muscle set was achieved as high as 99%. The optimal feature set consisted features such as Auto Regressive coefficients, Willison Amplitude and Slope Sign Change. The mean classification accuracy for different subjects, achieved using SVM and RF was 94.5% and 91.7% respectively.

Objectives. To optimize the absorbed organ dose in relation to the field of view for temporomandibular joint examinations in four cone beam computed tomography devices. Methods. An anthropomorphic adult head and neck phantom, and 192 LiF dosimeters (TLD-100) were used. The dosimeters were placed in the region corresponding to the lens, parotid glands, submandibular glands, and thyroid. Small, medium and large FOVs were selected on Orthopantomograph OP300 Maxio, PaX-i3D Smart, ORTHOPHOS XG, and i-CAT Next Generation device when it was possible. Results. A wide range of absorbed dose values was recorded for all organs due to the different exposure parameters of each device. The radiosensitive organ with the highest dose was the parotid glands. The devices with 5 × 5 cm FOV recorded a lower dose in this protocol, while for the device without a small FOV (≤5 × 5 cm), the lowest dose was observed with the large FOV (6 × 16 cm). Conclusions. We recommend a double exposure with an FOV of 5 × 5 cm in the OP300 Maxio, PaX-i3D Smart, and ORTHOPHOS XG device, while in the i-CAT Next Generation device, a single exposure FOV of 6 × 16 cm is indicated.

Melanoma is one of the most aggressive skin cancers. However, there remain many limitations in the current clinical treatments of it. Zinc oxide nanoparticles (ZnO NPs) have been considered to be a promising antitumor drug due to their excellent biocompatibility, biodegradability and biofunctionality. In this study, we prepared spherical ZnO NPs with an average diameter of less than 10 nm by a simple chemical method. According to the in vitro cytotoxicity assay, ZnO NPs in a certain concentration range (20–35 μg ml⁻¹) showed significant cytotoxicity to B16F10 melanoma cells, while having little effect on the viability of 3T3L1 fibroblasts. When cultured with B16F10 melanoma cells, ZnO NPs induced the generation of reactive oxygen and mitochondrial superoxide through the release of Zn²⁺, leading to oxidative stress in the cells, further reducing the mitochondrial membrane potential and decreasing the number of mitochondrial cristae. Furthermore, damaged mitochondria induced the release of apoptosis factors to promote cell apoptosis. FITC-Annexin V/propidium iodide double staining assay was used to analyze different apoptosis stages of B16F10 cells induced by ZnO NPs. A polymer hydrogel (Gel-F127-ZnO NPs) with Pluronic F127 as the carrier of ZnO NPs was fabricated for evaluating the antitumor effect of ZnO NPs in vivo. The in vivo experiment indicated that the tumor recurrence was significantly inhibited in tumor-bearing mice after treated with Gel-F127-ZnO NPs. Conclusively, ZnO NPs showed a strong antitumor effect both in vitro and in vivo.

Styrene-ethylene/butylene-styrene (SEBS) copolymer-in-mineral oil gel is an appropriate tissue-mimicking material to manufacture stable phantoms for ultrasound and photoacoustic imaging. Glycerol dispersion has been proposed to further tune the acoustic properties and to incorporate hydrophilic additives into SEBS gel. However, this type of material has not been investigated to produce wall-less vascular flow phantom for these imaging modalities. In this paper, the development of a wall-less vascular phantom for ultrasound and photoacoustic imaging is reported. Mixtures of glycerol/TiO₂-in-SEBS gel samples were manufactured at different proportions of glycerol (10%, 15%, and 20%) and TiO₂ (0% to 0.5%) to characterize their optical and acoustic properties. Optical absorption in the 500–950 nm range was independent of the amount of glycerol and TiO₂, while optical scattering increased linearly with the concentration of TiO₂. Acoustic attenuation and speed of sound were not influenced by the presence of TiO₂. The sample manufactured using weight percentages of 10% SEBS, 15% glycerol, and 0.2% TiO₂ was selected to make the vascular phantom. The phantom proved to be stable during the pulsatile blood-mimicking fluid (BMF) flow, without any observed damage to its structure or leaks. Ultrasound color Doppler images showed a typical laminar flow, while the B-mode images showed a homogeneous speckled pattern due to the presence of the glycerol droplets in the gel. The photoacoustic images of the phantom showed a well-defined signal coming from the surface of the phantom and from the vessels where BMF was flowing. The Spearman's correlations between the photoacoustic and tabulated spectra calculated from the regions containing BMF, in this case a mixture of salt solutions (NiCl₂ and CuSO₄), were higher than 0.95. Our results demonstrated that glycerol-in-SEBS gel was an adequate material to make a stable vascular flow phantom for ultrasound photoacoustic imaging.

Objectives. Volumetric modulated arc therapy (VMAT) allows for reduction of organs at risk (OAR) volumes receiving higher doses, but increases OAR volumes receiving lower radiation doses and can subsequently increasing associated toxicity. Therefore, reduction of this low-dose-bath is crucial. This study investigates personalizing the optimization of VMAT arc parameters (gantry start and stop angles) to decrease OAR doses. Materials and Methods. Twenty previously treated locally advanced non-small cell lung cancer (NSCLC) patients treated with half-arcs were randomly selected from our database. These plans were re-optimized with seven different arcs parameters; optimization objectives were kept constant for all plans. All resulting plans were reviewed by two clinicians and the optimal plan (lowest OAR doses and adequate target coverage) was selected. Furthermore, knowledge-based planning (KBP) model was developed using these plans as 'training data' to predict optimal arc parameters for individual patients based on their anatomy. Treatment plan complexity scores and deliverability measurements were performed for both optimal and original clinical plans. Results. The results show that different arc geometries resulted in different dose distributions to the OAR but target coverage was mostly similar. Different arc geometries were required for different patients to minimize OAR doses. Comparison of the personalized against the standard (2 half-arcs) plans showed a significant reduction in lung V₅ (lung volume receiving 5 Gy), mean lung dose and mean heart doses. Reduction in lung V₂₀ and heart V₃₀ were statistically insignificant. Plan complexity and deliverability measurements show the test plans can be delivered as planned. Conclusions. Our study demonstrated that personalizing arc parameters based on an individual patient's anatomy significantly reduces both lung and heart doses. Dose reduction is expected to reduce toxicity and improve the quality of life for these patients.

Application of multi-pinhole collimator in pinhole-based SPECT increases detection sensitivity. The presence of multiplexing in projection images due to the usage of multiple pinholes can further improve the sensitivity at the cost of adding data ambiguity. We are developing a next-generation adaptive brain-dedicated SPECT system –AdaptiSPECT-C. The AdaptiSPECT-C can adapt the multiplexing level and system sensitivity using adaptable pinhole modules. In this study, we investigated the performance of 4 data acquisition schemes with different multiplexing levels and sensitivities on cerebral SPECT imaging. Schemes #1, #2, and #3 have <1%, 67%, and 31% overall multiplexing, respectively, while the 4th scheme without multiplexing is considered as ground truth. The ground-truth and schemes #1–3 have 1.0, 1.7, 5.1, and 4.0 times higher sensitivity, respectively, compared to a dual-headed parallel-hole SPECT system at matched spatial resolution. A customized XCAT brain perfusion digital phantom emulating the distribution of I-123 N-isopropyl iodoamphetamine (IMP) in a 99th percentile size male was used for simulations. Data acquisition for each scheme was performed at two count levels (low-count and high-count relative to the recommended clinical count level). The normalized root-mean-square error (NRMSE) for schemes #1, #2, and #3 with the low-count (high-count) scenario showed 11%, 4%, and 5% (10%, 5%, and 6%) deviation, respectively, from that of the multiplex-free ground truth. For both the low-count and high-count scenarios, scheme #1 resulted in the least accurate activity ratio (AR) for almost all the analyzed gray-matter brain regions. Further schemes #2 or #3 led to the most accurate AR values with both low-count and high-count scenarios for all the analyzed gray-matter regions. It was thus observed that even with this large head size which leads to significant multiplexing levels, the higher sensitivity from multiplexing could to some extent mitigate the data ambiguity and be translated into reconstructed images of higher quality.

Deep learning has gained much popularity in solving challenging machine learning problems related to image, speech classification, etc. Research has been conducted to apply deep learning models in emotion classification based on physiological signals such as EEG. Most of the research works have based their model on the spatial aspects of the EEG. However, the emotion features in EEG are spread across the time domain during an emotional episode. Therefore, in this work, the emotion classification problem is modelled as a sequence classification problem. The power band frequency based features of every time segment of EEG sequences generated from 32-channel EEG data are used to train three different models of Long Short-Term Memory (LSTM1, LSTM2, and LSTM3). Four class (HVHA, HVLA, LVHA, and LVLA) classification experiments were performed based on the valence and arousal emotion models. The LSTM3 model with 128 memory cells achieved the highest classification accuracy of 90%, whereas LSTM1 (32 cells) and LSTM2 (64 cells) yielded classification accuracies of 85% and 89% respectively. Further, the impact of segment size on classification accuracy was also investigated in this work. Results obtained indicate that a smaller segment size leads to higher classification accuracy using LSTM models.

Purpose. To investigate image intensity histograms as a potential source of useful imaging biomarkers in both a clinical example of detecting immune-related colitis (irColitis) in ¹⁸F-FDG PET/CT images of immunotherapy patients and an idealized case of classifying digital reference objects (DRO). Methods. Retrospective analysis of bowel ¹⁸F-FDG uptake in N = 40 patients receiving immune checkpoint inhibitors was conducted. A CNN trained to segment the bowel was used to generate the histogram of bowel ¹⁸F-FDG uptake, and percentiles of the histogram were considered as potential metrics for detecting inflammation associated with irColitis. A model of the colon was also considered using cylindrical DRO. Classification of DRO with different intensity distributions was undertaken under varying geometry and noise settings. Results. The most predictive biomarker of irColitis was the 95th percentile of the bowel SUV histogram (SUV_95%). Patients later diagnosed with irColitis had a significantly higher increase in SUV_95% from baseline to first on-treatment PET than patients who did not experience irColitis (p = 0.02). An increase in SUV_95%> + 40% separated pre-irColitis change from normal variability with a sensitivity of 75% and specificity of 88%. Furthermore, histogram percentiles were ideal metrics for classifying 'hot center' and 'cold center' DRO, and were robust to varying DRO geometry and noise, and to the presence of spoiler volumes unrelated to the detection task. Conclusions. The 95th percentile of the bowel SUV histogram was the optimal metric for detecting irColitis on ¹⁸F-FDG PET/CT. Image intensity histograms are a promising source of imaging biomarkers for clinical tasks.

The development of foot ulcers is a common consequence of severe diabetes. Due to vascular disorders and impeded healing caused by the disease, most foot ulcers have been reported to be affected by body weight and progress with time. Also, abnormal distribution of plantar pressures has been observed to cause the formation of additional ulcers, which may collectively lead to traumatic amputations. While a study of such pathophysiology is not possible through experiments, a few computational modelling works have investigated diabetic foot ulcers. To date, ulcers with a few sizes and locations have been studied, and their effect on the plantar stresses has been quantified. In this work, we have attempted to study the effect of all possible ulcer locations on the generated plantar peak stresses and peak stress locations where additional ulcers may form. Also, the effect of ulcer location on the possible ulcer growth was investigated. A full-scale foot model was developed and a total of 52 ulcer locations were simulated separately, with standing and walking loads. The generated stresses were normalised with the foot size and statistically analysed to develop novel formulations for predicting peak plantar stresses and their locations for any known ulcer location. The results from this study are anticipated to provide important guidelines to doctors and medical practitioners for predicting foot ulcer progression in diabetic patients with existing ulcers and allow the administration of timely preventive interventions.

Objective. Extraction of temporal features of neuronal activity from electrophysiological data can be used for accurate classification of neural networks in healthy and pathologically perturbed conditions. In this study, we provide an extensive approach for the classification of human in vitro neural networks with and without an underlying pathology, from electrophysiological recordings obtained using a microelectrode array (MEA) platform. Approach. We developed a Dirichlet mixture (DM) Point Process statistical model able to extract temporal features related to neurons. We then applied a machine learning algorithm to discriminate between healthy control and pathologically perturbed in vitro neural networks. Main Results. We found a high degree of separability between the classes using DM point process features (p-value <0.001 for all the features, paired t-test), which reaches 93.10 of accuracy (92.37 of ROC AUC) with the Random Forest classifier. In particular, results show a higher latency in firing for pathologically perturbed neurons (43 ± 16 ms versus 67 ± 31 ms, ${\mu }_{IG}$ feature distribution). Significance. Our approach has been successful in extracting temporal features related to the neurons' behaviour, as well as distinguishing healthy from pathologically perturbed networks, including classification of responses to a transient induced perturbation.

The use of energy-based devices to treat genitourinary syndrome of menopause, termed vaginal thermotherapy (VTT), has gained significant interest in recent years. Among the primary safety concerns of this relatively new procedure is the possibility of unintentionally heating tissues adjacent to the vaginal wall, i.e., heating too deeply. Herein we use numerical simulations to evaluate monopolar radiofrequency-based (RF) VTT specifically focusing on the resultant depth of heating through a range of input parameters. Varying RF power, exposure time, and the simulated rate of blood perfusion, we map the parameter space identifying which combinations of input parameters are likely to heat past the depth of the vaginal wall and affect adjacent tissue. We found that the device parameters commonly used in the literature are likely to heat past the vaginal wall and merit further investigation. In addition, we found that the parameter typically used to describe VTT devices, total energy delivered, does not reliably indicate the resultant depth of heat dispersion.

The recent development of deep learning approaches has revoluted medical data processing, including semantic segmentation, by dramatically improving performance. Automated segmentation can assist radiotherapy treatment planning by saving manual contouring efforts and reducing intra-observer and inter-observer variations. However, training effective deep learning models usually Requires a large amount of high-quality labeled data, often costly to collect. We developed a novel semi-supervised adversarial deep learning approach for 3D pelvic CT image semantic segmentation. Unlike supervised deep learning methods, the new approach can utilize both annotated and un-annotated data for training. It generates un-annotated synthetic data by a data augmentation scheme using generative adversarial networks (GANs). We applied the new approach to segmenting multiple organs in male pelvic CT images. CT images without annotations and GAN-synthesized un-annotated images were used in semi-supervised learning. Experimental results, evaluated by three metrics (Dice similarity coefficient, average Hausdorff distance, and average surface Hausdorff distance), showed that the new method achieved comparable performance with substantially fewer annotated images or better performance with the same amount of annotated data, outperforming the existing state-of-the-art methods.

In view of efficiency, simple operation, and affordable cost and disposability, quartz tuning fork systems form good candidates for mechanical-based biosensors in point of care applications. Based on the geometrical structure, the frequency response of the tuning fork- based sensors is dependent on the location of absorbed samples. In order to have the maximum efficiency and sensitivity, the optimized condition of sample loading on the fork structures should be considered. In this regard, here, we have determined the optimized sample location to be on the prongs of the quartz tuning fork by calculating the frequency response of the quartz tuning fork using the finite element method. From an application point of view, we have obtained an agreement between the calculational method and the experimental excitation technique of the structure. The results from our study show that by using an appropriate location for the sample, the quartz tuning fork could be exploited with high sensitivity.

Tissue mimicking phantom materials with thermal and dielectric equivalence are vital for the development of microwave diagnostics and treatment. The current phantoms representing fat tissue are challenged by mechanical integrity at relevant temperatures coupled with complex production protocols. We have employed two types of nanocellulose (cellulose nanocrystals and oxidized cellulose nanocrystals) as reinforcement in gelatin stabilized emulsions for mimicking fat tissue. The nanocellulose-gelatin stabilized emulsions were evaluated for their dielectric properties, the moduli-temperature dependence using small deformation rheology, stress-strain behavior using large deformation, and their compliance to quality assurance guidelines for superficial hyperthermia. All emulsions had low permittivity and conductivity within the lower microwave frequency band, accompanied by fat equivalent thermal properties. Small deformation rheology showed reduced temperature dependence of the moduli upon addition of nanocellulose, independent of type. The cellulose nanocrystals gelatin reinforced emulsion complied with the quality assurance guidelines. Hence, we demonstrate that the addition of cellulose nanocrystals to gelatin stabilized emulsions has the potential to be used as fat phantoms for the development of microwave diagnostics and treatment.

Electrocardiograms (ECG) recorded from everyday objects, such as wearables, fitness machines or smart steering wheels are becoming increasingly common. Applications are diverse and include health monitoring, athletic performance optimization, identification, authentication, and entertainment. In this study we report the design and implementation of an innovative ECG simulator, providing simulation of signal related artifacts and a dynamically adjustable skin-electrode interface model. The ECG simulator includes a unique combination of features: emulation of time dependent skin-electrode impedance, adjustable differential and common-mode interference, generation of lead-off events and analog front-end output digitalization. The skin-electrode capacitance range is 1 nF-255 nF and the resistance span is 4 kΩ-996 kΩ. System's functionality is demonstrated using a commercially available ECG front-end. The simulated SNR degradation introduced by the ECG simulator is under 0.1 dB. Results show that the skin-electrode interface can have a significant impact in the acquired waveforms. Impedance electrode imbalance, specifically of the resistive component, can generate artifacts which can be misinterpreted has arrhythmias. The proposed device can be useful for hardware and software ECG development and for training physicians and nurses to readily recognize skin-electrode impedance related artifacts.

Electroencephalogram (EEG) signals are crucial to Brain-Computer Interfacing (BCI). However, these are vulnerable to a variety of unintended artifacts that could negatively impact the precise brain function assessment. This paper provides a new algorithm to eliminate Electrical Shift and Linear Trend artifact (ESLT) in EEG using Singular Spectrum Analysis (SSA) and Enhanced local Polynomial (LP) Approximation-based Total Variation (EPATV). The contaminated single channel EEG is subdivided into multiple bands of frequency components by SSA. In order to acquire all LP and TV components, EPATV filtering is applied over the contaminated component frequency band. Filtered sub-signal is collected by subtracting both the LP and TV components from the component contaminated frequency band. Then, the addition of filtered sub-signal and remaining SSA frequency band components yield the final denoised EEG signal. The effectiveness of the proposed method in this paper is evaluated using the data obtained from three databases and compared with the existing methods. From the extensive simulation results, it is inferred that the algorithm discussed in the paper is effective when compared the existing methods, exhibiting a highest averaged Correlation Coefficient (CC) of 0.9534, averaged Signal to Noise Ratio (SNR) of 10.2208dB, lowest averaged Relative Root Mean Square Error (RRMSE) value 0.2787 and averaged Mean absolute Error (MAE) in α band value of 0.0557. The algorithm presented in this paper may be a viable choice for extracting ESLT artifact from a small streaming section of the EEG without requirement of the initial calibration or enormous EEG data.

Objective. Electrical stimulation of the auricular vagus nerve is a non-invasive neuromodulation technique that has been used for various conditions, including depression, epilepsy, headaches, and cerebral ischemia. However, unwanted non-vagal nerve stimulations can occur because of diffused stimulations. The objective of this study is to develop a region-specific non-invasive vagus nerve stimulation (VNS) technique using the millimeter wave (MMW) as a stimulus for the auricular branch of the vagus nerve (ABVN). Approach. A numerical simulation was conducted to ascertain whether the MMW could excite the ABVN in the human outer-ear with a millimeter-scale spatial resolution. Additionally, MMW-induced neuronal responses in seven mice were evaluated. Transcutaneous auricular VNS (ta-VNS) was applied to the cymba conchae innervated by the AVBN using a 60-GHz continuous wave (CW). As a control, the auricle's exterior margin was stimulated and referred to as transcutaneous auricular non-vagus nerve stimulation (ta-nonVNS). During stimulation, the local field potential (LFP) in the nucleus tractus solitarii (NTS), an afferent vagal projection site, was recorded simultaneously. Main results. The ta-VNS with a stimulus level of 13 dBm showed a significant increase in the LFP power in the NTS. The mean increases in power (n = 7) in the gamma high and gamma very high bands were 8.6 ± 2.0% and 18.2 ± 5.9%, respectively. However, the ta-nonVNS with a stimulus level of 13 dBm showed a significant decrease in the LFP power in the NTS. The mean decreases in power in the beta and gamma low bands were 11.0 ± 4.4% and 10.8 ± 2.8%, respectively. These findings suggested that MMW stimulation clearly induced a different response according to the presence of ABVN. Significance. Selective auricular VNS is feasible using the MMW. This study provides the basis for the development of a new clinical treatment option using the stimulation of the ta-VNS with a square millimeter spatial resolution.

Numerous diseases alter the esophagus elasticity, such as eosinophilic esophagitis and esophageal motility disorders like achalasia. The possibility to measure these modifications using minimally invasive techniques is a key issue for the diagnosis of such pathologies. The commercially available Endoflip^TM (endoluminal functional lumen imaging probe) can be used to measure the luminal cross-sectional diameter of the esophagus at different points and over time, and is used in clinical routine to assess esophageal distensibility. We used this probe to track the propagation of shear waves similar to those that are produced naturally by natural waves, to compute wavelength of the esophagus using passive elastography algorithms. To assess the feasibility of such measurements, we compared the wavelengths obtained with the probe in polyvinyl alcohol (PVA) gel tubes to those obtained for the same tubes with optical tracking of their edges using a camera. We first compared the wavelength obtained with homogeneous gel tubes with both techniques, and then used paired gel tubes of different elasticities to investigate the possibility to measure different wavelengths. Although, the wavelength computed using the probe and the camera showed some small differences, qualitative differentiation of the tubes was achieved when using paired tubes with different elasticities. Using the camera, a wavelength of 61 mm was measured for the hard tube, and 35 mm for the soft tube. Using the probe, wavelengths of 61 mm and 38 mm were measured, respectively. Therefore, we demonstrate here the feasibility of using this probe to track wave propagation and to determine the wavelengths in gel tubes of different stiffnesses. This analysis was also taken to a preliminary in-vivo study that allowed tracking of natural waves in the esophagus using the luminal probe, which indicates that this technique can also be used in vivo to measure the stiffness of the esophagus.

One of the prominent reasons behind the deterioration of cardiovascular conditions is hypertension. Due to lack of specific symptoms, sometimes existing hypertension goes unnoticed until significant damage happens to the heart or any other body organ. Monitoring of BP at a higher frequency is necessary so that we can take early preventive measures to control and keep it within the normal range. The cuff-based method of measuring BP is inconvenient for frequent daily measurements. The cuffless BP measurement method proposed in this paper uses features extracted from the electrocardiogram (ECG) and photoplethysmography (PPG). ECG and PPG both have distinct characteristics, which change with the change of blood pressure levels. Feature extraction and hybrid feature selection algorithms are followed by a generalized penalty-based regression technique led to a new BP measurement process that uses the minimum number of features. The performance of the proposed technique to measure blood pressure was compared to an approach using an ordinary linear regression method with no feature selection and to other contemporary techniques. MIMIC-II database was used to train and test our proposed method. The root mean square error (RMSE) for systolic blood pressure (SBP) improved from 11.2 mmHg to 5.6 mmHg when the proposed technique was implemented and for diastolic blood pressure (DBP) improved from 12.7 mmHg to 6.69 mmHg. The mean absolute error (MAE) was found to be 4.91 mmHg for SBP and 5.77 mmHg for DBP, which have shown improvement over other existing cuffless techniques where the substantial number of patients, as well as feature selection algorithm, were implemented. In addition, according to the British Hypertension Society standard (BHS) standard for cuff-based BP measurement, the criteria for acceptable measurement are to achieve at least grade B; our proposed method also satisfies this criterion.

In the medical field, automated and computerised analytic tools are essential for faster disease diagnosis. The main objective of this research work is to classify the leukocytes accurately into four different subtypes based on the pattern of the nucleus. The features are extracted from the segmented nucleus, which play a vital role in the pattern recognition. The technique comprises a novel idea of computing the statistical measures such as peak difference and standard deviation of the radon transformed graph for a single angle of rotation along with other features. Three Gray Level Co-occurrence Matrix (GLCM) based features, two geometric features and four RST moment invariants are also extracted for feature fusion. The fused feature vectors are trained and evaluated using random forest classification algorithm.This method provides an overall accuracy of 97.61% and it is able to determine the lymphocyte, neutrophil and eosinophil with 100% accuracy. The classification without incorporating radon transform features is also performed which provides an accuracy of only 80.95%.

In this work, we study the measurement of blood velocity with contrast enhanced computed tomography. The reconstruction is based on CT projections perpendicular to the main axis of the vessel and on a partial differential equation describing the propagation of the contrast agent. The inverse problem is formulated as an optimal control problem with the transport equation as constraint. The velocity field is obtained with stationary and unstationary Navier–Stokes equations and it is reconstructed with the adjoint method. The velocity and the density of the contrast agent are well reconstructed. The reconstruction results obtained are better for the axial component of the velocity than for transverse components.

Objective. Finite element method (FEM) simulations of the electric field magnitude (EF) are commonly used to estimate the affected tissue surrounding the active contact of deep brain stimulation (DBS) leads. Previous studies have found that DBS starts to noticeably activate axons at approximately 0.2 V mm⁻¹, corresponding to activation of 3.4 μm axons in simulations of individual axon triggering. Most axons in the brain are considerably smaller however, and the effect of the electric field is thus expected to be stronger with increasing EF as more and more axons become activated. The objective of this study is to estimate the fraction of activated axons as a function of electric field magnitude. Approach. The EF thresholds required for axon stimulation of myelinated axon diameters between 1 and 5 μm were obtained from a combined cable and Hodgkin-Huxley model in a FEM-simulated electric field from a Medtronic 3389 lead. These thresholds were compared with the average axon diameter distribution from literature from several structures in the human brain to obtain an estimate of the fraction of axons activated at EF levels between 0.1 and 1.8 V mm⁻¹. Main results. The effect of DBS is estimated to be 47·EF—8.8% starting at a threshold level EF_t0 = 0.19 V mm⁻¹. Significance. The fraction of activated axons from DBS in a voxel is estimated to increase linearly with EF above the threshold level of 0.19 V mm⁻¹. This means linear regression between EF above 0.19 V mm⁻¹ and clinical outcome is a suitable statistical method when doing improvement maps for DBS.

Axial Vertebral Rotation (AVR) is a significant indicator of adolescent idiopathic scoliosis (AIS). A host of methods are provided to measure AVR on coronal plane radiographs or 3D vertebral model. This paper provides a method of automatic AVR measurement in 3D vertebral model that is based on point cloud segmentation neural network and the tip of the spinous process searching algorithm. An improved PointNet using multi-input and attention mechanism named Multi-Input PointNet is proposed, which can segment the upper and lower endplates of the vertebral model accurately to determine the transverse plane of vertebral model. An algorithm is developed to search the tip of the spinous process according to the special structure of vertebrae. AVR angle is measured automatically using the midline of vertebral model and projection of y-axis on the transverse plane of vertebral model based on points obtained above. We compare automatic measurement results with manual measurement results on different vertebral models. The experiment shows that automatic results can achieve accuracy of manual measurement results and the correlation coefficient of them is 0.986, proving our automatic AVR measurement method performs well.

Super-resolution ultrasound (SR-US) imaging allows visualization of microvascular structures as small as tens of micrometers in diameter. However, use in the clinical setting has been impeded in part by ultrasound (US) acquisition times exceeding a breath-hold and by the need for extensive offline computation. Deep learning techniques have been shown to be effective in modeling the two more computationally intensive steps of microbubble (MB) contrast agent detection and localization. Performance gains by deep networks over conventional methods are more than two orders of magnitude and in addition the networks can localize overlapping MBs. The ability to separate overlapping MBs allows use of higher contrast agent concentrations and reduces US image acquisition time. Herein we propose a fully convolutional neural network (CNN) architecture to perform the operations of MB detection as well as localization in a single model. Termed SRUSnet, the network is based on the MobileNetV3 architecture modified for 3-D input data, minimal convergence time, and high-resolution data output using a flexible regression head. Also, we propose to combine linear B-mode US imaging and nonlinear contrast pulse sequencing (CPS) which has been shown to increase MB detection and further reduce the US image acquisition time. The network was trained with in silico data and tested on in vitro data from a tissue-mimicking flow phantom, and on in vivo data from the rat hind limb (N = 3). Images were collected with a programmable US system (Vantage 256, Verasonics Inc., Kirkland, WA) using an L11–4v linear array transducer. The network exceeded 99.9% detection accuracy on in silico data. The average localization accuracy was smaller than the resolution of a pixel (i.e. ${\boldsymbol{\lambda }}/8$). The average processing time on a Nvidia GeForce 2080Ti GPU was 64.5 ms for a 128 × 128-pixel image.

The purpose of this work was to investigate the use of the Varian Portal Dosimetry application in conjunction with in vivo megavoltage portal images on a Varian Halcyon O-ring type linear accelerator as an in vivo dosimetry constancy (IVDc) tool for pelvis and head/neck patients receiving VMAT treatments. Sensitivity testing was conducted on phantoms with varying thicknesses (0.2 cm–1.0 cm) using static and modulated fields. A cohort of 96 portal dose images across eight patients was then compared with PTV metrics derived from daily CBCT image based treatment plan re-calculations to determine whether the IVDc tool could detect gross inter-fraction anatomical changes. A final cohort of 315 portal dose images across 22 patients was then assessed to demonstrate the application of IVDc tool. The IVDc tool, using 2%/2 mm criteria, detected all phantom thickness changes of 1.0 cm, some phantom thickness changes of 0.5 cm, and no changes of 0.2 cm. For the cohort of 96 results, a IVDc passing criteria of 95% (2%, 2 mm) was able to identify all cases that had PTV metric changes of 2% or more. Using the IVDc tool on the cohort of 315 results, and the IVDc passing criteria of 95%, resulted in 74 IVDc failures. A simple, easy to implement, methodology has been presented that is capable of detecting gross inter-fraction changes in patient geometry on the Varian Halcyon O-ring linac linear accelerator.

This study aimed to dosimetrically compare and evaluate the flattening filter-free (FFF) photon beam-based three-dimensional conformal radiotherapy (3DCRT), intensity-modulated radiation therapy (IMRT), and volumetric modulated arc therapy (VMAT) for lung stereotactic body radiotherapy (SBRT). RANDO phantom computed tomography (CT) images were used for treatment planning. Gross tumor volumes (GTVs) were delineated in the central and peripheral lung locations. Planning target volumes (PTVs) was determined by adding a 5 mm margin to the GTV. 3DCRT, IMRT, and VMAT plans were generated using a 6-MV FFF photon beam. Dose calculations for all plans were performed using the anisotropic analytical algorithm (AAA) and Acuros XB algorithms. The accuracy of the algorithms was validated using the dose measured in a CIRS thorax phantom. The conformity index (CI), high dose volume (HDV), low dose location (D_2cm), and homogeneity index (HI) improved with FFF-VMAT compared to FFF-IMRT and FFF-3DCRT, while low dose volume (R_50%) and gradient index (GI) showed improvement with FFF-3DCRT. Compared with FFF-3DCRT, a drastic decrease in the mean treatment time (TT) value was observed with FFF-VMAT for different lung sites between 57.09% and 60.39%, while with FFF-IMRT it increased between 10.78% and 17.49%. The dose calculation with Acuros XB was found to be superior to that of AAA. Based on the comparison of dosimetric indices in this study, FFF-VMAT provides a superior treatment plan to FFF-IMRT and FFF-3DCRT in the treatment of peripheral and central lung PTVs. This study suggests that Acuros XB is a more accurate algorithm than AAA in the lung region.

Microwave ablation is under investigation as a minimally-invasive treatment for uterine fibroids. Computational models play a vital role in the development, evaluation and characterization of candidate ablation devices. The temperature-dependent dielectric properties of fibroid tissue are essential for accurate computational modeling. Objective: To measure the broadband temperature-dependent dielectric properties of uterine fibroids excised during hysterectomy procedures. Methods: The open-ended coaxial probe method was employed for measuring the broadband dielectric properties of freshly excised human uterine fibroid samples (n = 6) obtained from an IRB-approved tissue bank. The dielectric properties (relative permittivity, ε_r, and effective electrical conductivity, σ_eff) were evaluated at temperatures ranging from 23 °C–150 °C, over the frequency range of 0.5–6 GHz. Linear piecewise parametrization with respect to temperature and quadratic parametrization with respect to frequency was applied to characterize broadband temperature-dependent dielectric properties of fibroid tissue. Results: The baseline room temperature values of ε_r vary from 57.5 ± 5.29 to 44.5 ± 5.77 units and σ_eff changes from 0.91 ± 0.19 to 6.02 ± 0.7 S m⁻¹ over the frequency range of 0.5–6 GHz. At temperatures close to the water vaporization point, ε_r, drops considerably i.e. to 12%–14% of its baseline value for all measured frequencies. σ_eff values initially rise till 98 °C and then fall to 11%–13% of their baseline values at 125 °C for frequencies ≤2.45 GHz. The σ_eff follows a decreasing trend for frequencies >2.45 GHz and drops to ∼6 % of their baseline room temperature values. Conclusion: The temperature dependent dielectric properties of uterine fibroid tissues over microwave frequency range are reported for the first time in this study. Parametric models of uterine fibroid dielectric properties are also presented for incorporation within computational models of microwave ablation of fibroids.

Purpose. To investigate indirect radiation-induced changes in airways as precursors to atelectasis post radiation therapy (RT). Methods. Three Wisconsin Miniature Swine (WMS^TM) underwent a research course of 60 Gy in 5 fractions delivered to a targeted airway/vessel in the inferior left lung. The right lung received a max point dose <5 Gy. Airway segmentation was performed on the pre- and three months post-RT maximum inhale phase of the four-dimensional (4D) computed tomography (CT) scans. Changes in luminal area (Ai) and square root of wall area ($\sqrt{{WA}}$) for each airway were investigated. Changes in ventilation were assessed using the Jacobian ratio and were measured in three different regions: the inferior left lung <5 Gy (ILL), the superior left lung <5 Gy (SLL), and the contralateral right lung <5 Gy (RL). Results. Airways (n = 25) in the right lung for all swine showed no significant changes (p = 0.48) in Ai post-RT compared to pre-RT. Airways (n = 28) in the left lung of all swine were found to have a significant decrease (p < 0.001) in Ai post-RT compared to pre-RT, correlated (Pearson R = −0.97) with airway dose. Additionally, $\sqrt{{WA}}$ decreased significantly (p < 0.001) with airway dose. Lastly, the Jacobian ratio of the ILL (0.883) was lower than that of the SLL (0.932) and the RL (0.955). Conclusions. This work shows that for the swine analyzed, there were significant correlations between Ai and $\sqrt{{WA}}$ change with radiation dose. Additionally, there was a decrease in lung function in the regions of the lung supplied by the irradiated airways compared to the regions supplied by unirradiated airways. These results support the hypothesis that airway dose should be considered during treatment planning in order to potentially preserve functional lung and reduce lung toxicities.

Kilovoltage cone-beam computed tomography (CBCT)-based image-guided radiation therapy (IGRT) is used for daily delivery of radiation therapy, especially for stereotactic body radiation therapy (SBRT), which imposes particularly high demands for setup accuracy. The clinical applications of CBCTs are constrained, however, by poor soft tissue contrast, image artifacts, and instability of Hounsfield unit (HU) values. Here, we propose a new deep learning-based method to generate synthetic CTs (sCT) from thoracic CBCTs. A deep-learning model which integrates histogram matching (HM) into a cycle-consistent adversarial network (Cycle-GAN) framework, called HM-Cycle-GAN, was trained to learn mapping between thoracic CBCTs and paired planning CTs. Perceptual supervision was adopted to minimize blurring of tissue interfaces. An informative maximizing loss was calculated by feeding CBCT into the HM-Cycle-GAN to evaluate the image histogram matching between the planning CTs and the sCTs. The proposed algorithm was evaluated using data from 20 SBRT patients who each received 5 fractions and therefore 5 thoracic CBCTs. To reduce the effect of anatomy mismatch, original CBCT images were pre-processed via deformable image registrations with the planning CT before being used in model training and result assessment. We used planning CTs as ground truth for the derived sCTs from the correspondent co-registered CBCTs. The mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and normalized cross-correlation (NCC) indices were adapted as evaluation metrics of the proposed algorithm. Assessments were done using Cycle-GAN as the benchmark. The average MAE, PSNR, and NCC of the sCTs generated by our method were 66.2 HU, 30.3 dB, and 0.95, respectively, over all CBCT fractions. Superior image quality and reduced noise and artifact severity were seen using the proposed method compared to the results from the standard Cycle-GAN method. Our method could therefore improve the accuracy of IGRT and corrected CBCTs could help improve online adaptive RT by offering better contouring accuracy and dose calculation.

In interior computed tomography (CT), the x-ray beam is collimated to a limited field-of-view (FOV) (e.g. the volume of the heart) to decrease exposure to adjacent organs, but the resulting image has a severe truncation artifact when reconstructed with traditional filtered back-projection (FBP) type algorithms. In some examinations, such as cardiac or dentomaxillofacial imaging, interior CT could be used to achieve further dose reductions. In this work, we describe a deep learning (DL) method to obtain artifact-free images from interior CT angiography. Our method employs the Pix2Pix generative adversarial network (GAN) in a two-stage process: (1) An extended sinogram is computed from a truncated sinogram with one GAN model, and (2) the FBP reconstruction obtained from that extended sinogram is used as an input to another GAN model that improves the quality of the interior reconstruction. Our double GAN (DGAN) model was trained with 10 000 truncated sinograms simulated from real computed tomography angiography slice images. Truncated sinograms (input) were used with original slice images (target) in training to yield an improved reconstruction (output). DGAN performance was compared with the adaptive de-truncation method, total variation regularization, and two reference DL methods: FBPConvNet, and U-Net-based sinogram extension (ES-UNet). Our DGAN method and ES-UNet yielded the best root-mean-squared error (RMSE) (0.03 ± 0.01), and structural similarity index (SSIM) (0.92 ± 0.02) values, and reference DL methods also yielded good results. Furthermore, we performed an extended FOV analysis by increasing the reconstruction area by 10% and 20%. In both cases, the DGAN approach yielded best results at RMSE (0.03 ± 0.01 and 0.04 ± 0.01 for the 10% and 20% cases, respectively), peak signal-to-noise ratio (PSNR) (30.5 ± 2.6 dB and 28.6 ± 2.6 dB), and SSIM (0.90 ± 0.02 and 0.87 ± 0.02). In conclusion, our method was able to not only reconstruct the interior region with improved image quality, but also extend the reconstructed FOV by 20%.

Synchrony Respiratory Tracking system adapted from CyberKnife has been introduced in Radixact to compensate the tumor motion caused by respiration. This study aims to compare the modeling accuracy of the Synchrony system between Radixact and CyberKnife. Two Synchrony plans based on fiducial phantoms were created for CyberKnife and Radixact, respectively. Different respiratory motion traces were used to drive a motion platform to move along the superoinferior and left-right direction. The cycle time and the amplitude of target/surrogate motion of one selected motion trace were scaled to investigate the dependence of modeling accuracy on the motion characteristic. The predicted target position, the correlation error, potential difference (Radixact only) and standard error (CyberKnife only) were extracted from raw data or log files of the two systems. The modeling accuracy was evaluated by calculating the root-mean-square (RMS) error between the predicted target positions and the input motion trace. A threshold T95 within which 95% of the potential difference or the standard error lay was defined and evaluated. Except for the motion trace with a small amplitude and a good (linear) correlation between target and surrogate motion, Radixact showed smaller RMS errors than CyberKnife. The RMS error of both systems increased with the motion amplitude and showed a decreasing trend with the increasing cycle time. No correlation was found between the RMS error and the amplitude of surrogate motion. T95 could be a good estimator of modeling accuracy for CyberKnife rather than Radixact. The correlation error defined in Radixact were largely affected by the number of fiducial markers and the setup error. In general, the modeling accuracy of the Radixact Synchrony system is better than that of the CyberKnife Synchrony system under unfavorable conditions.

The purpose of this study was to develop and evaluate a framework to support automated standardized testing and analysis of Cone Beam Computed Tomography (CBCT) image quality QA across multiple institutions. A survey was conducted among the participating institutions to understand the variability of the CBCT QA practices. A commercial, automated software platform was validated by seven institutions participating in a consortium dedicated to automated quality assurance. The CBCT image analysis framework was used to compare periodic QA results among 23 linear accelerators (linacs) from seven institutions. The CBCT image quality metrics (geometric distortion, spatial resolution, contrast, HU constancy, uniformity and noise) data are plotted as a function of means with the upper and lower control limits compared to the linac acceptance criteria and AAPM recommendations. For example, mean geometric distortion and HU constancy metrics were found to be 0.13 mm (TG142 recommendation: ≤2 mm) and 13.4 respectively (manufacturer acceptance specification: ≤±50).Image upload and analysis process was fully automated using a MATLAB-based platform. This analysis enabled a quantitative, longitudinal assessment of the performance of quality metrics which were also compared across 23 linacs. For key CBCT parameters such as uniformity, contrast, and HU constancy, all seven institutions used stricter goals than what would be recommended based on the analysis of the upper and lower control limits. These institutional goals were also found to be stricter than that found in AAPM published guidance. This work provides a reference that could be used to machine-specific optimized tolerance of CBCT image maintenance via control charts to monitor performance we well as the sensitivity of different tests in support of a broader quality assurance program. To ensure the daily image quality needed for patient care, the optimized statistical QA metrics recommended to using along with risk-based QA.

High-energy medical linear accelerator (Linac) has been widely used for treating cancer patients. However, with its effectiveness, high-energy linac yields an undesirable amount of neutron contamination. An MCNPX code version 2.6.0 was used for calculating photoneutron contamination from Varian Clinac iX 15 MV linac heads in this study. The fast neutrons were dominantly produced inside the linac head. The neutron fluence, absorbed dose, and dose equivalent calculations occurred inside a linac head and a water phantom model. The fast neutrons begin to be moderated after 1 cm inside the water phantom by calculating the energy spectra. Variations in the field sizes from 2 × 2, 5 × 5, 10 × 10, and 15 × 15 cm² show that the neutron production yield would increase for larger field sizes. The maximum neutron dose equivalents are 3.745; 7.687; 11.794 and 14.197 μSv/MU for 2 × 2, 5 × 5, 10 × 10 and 15 × 15 cm² field sizes, respectively. These calculations predict the photoneutron characteristics with more detail inside a treated patient during radiation therapy procedures.

Quality Control (QC) tests in mammography are very important, since mammograms have been used as a population-based screening test for more than 30 years and QCs lead to better image quality and radiation safety for patients. European guidelines, EUREF and EFOMP protocols provide comprehensive QC guidelines for digital mammography (DM) and digital breast tomosynthesis (DBT) units, respectively. We developed a novel, fast, free and platform independent software (named MAMMO_QC) for QC performance tests in DM and tomosynthesis, based on the aforementioned guidelines. MAMMO_QC consists of a series of ImageJ plugins for DM and DBT. It does not require any programming knowledge and can be used to evaluate several performance parameters, such as pre-sampled modulation transfer function (pMTF), normalised noise power spectrum (NNPS), detective quantum efficiency (DQE), contrast-detail analysis based on the CDMAM 3.4 and 4.0 test tools, homogeneity and artefacts, automatic exposure control (AEC) performance. The user can use MAMMO_QC for acceptance, commissioning and routine QC performance analysis based on the European guidelines. We validated our results against well-established software products used in mammography and DBT (i.e., COQ, OBJ_IQ_reduced and Artinis CDMAM Analyzer). All the average relative differences were within 5.5%, and several years of usage and testing allows us to consider MAMMO_QC as an accurate and reliable tool for QC on DM and DBT systems. Our developed software for DM and DBT computes almost all the parameters stated in the European, EUREF and EFOMP guidelines. To the best of our knowledge, no such software has been developed so far.