Machine learning-based diagnosis and risk classification of coronary artery disease using myocardial perfusion imaging SPECT: A radiomics study.

This study aimed to investigate the diagnostic performance of machine learning-based radiomics analysis to diagnose coronary artery disease status and risk from rest/stress Myocardial Perfusion Imaging (MPI) single-photon emission computed tomography (SPECT). A total of 395 patients suspicious of coronary artery disease who underwent 2-day stress-rest protocol MPI SPECT were enrolled in this study. The left ventricle myocardium, excluding the cardiac cavity, was manually delineated on rest and stress images to define a volume of interest. Added to clinical features (age, sex, family history, diabetes status, smoking, and ejection fraction), a total of 118 radiomics features, were extracted from rest and stress MPI SPECT images to establish different feature sets, including Rest-, Stress-, Delta-, and Combined-radiomics (all together) feature sets. The data were randomly divided into 80% and 20% subsets for training and testing, respectively. The performance of classifiers built from combinations of three feature selections, and nine machine learning algorithms was evaluated for two different diagnostic tasks, including 1) normal/abnormal (no CAD vs. CAD) classification, and 2) low-risk/high-risk CAD classification. Different metrics, including the area under the ROC curve (AUC), accuracy (ACC), sensitivity (SEN), and specificity (SPE), were reported for models' evaluation. Overall, models built on the Stress feature set (compared to other feature sets), and models to diagnose the second task (compared to task 1 models) revealed better performance. The Stress-mRMR-KNN (feature set-feature selection-classifier) reached the highest performance for task 1 with AUC, ACC, SEN, and SPE equal to 0.61, 0.63, 0.64, and 0.6, respectively. The Stress-Boruta-GB model achieved the highest performance for task 2 with AUC, ACC, SEN, and SPE of 0.79, 0.76, 0.75, and 0.76, respectively. Diabetes status from the clinical feature family, and dependence count non-uniformity normalized, from the NGLDM family, which is representative of non-uniformity in the region of interest were the most frequently selected features from stress feature set for CAD risk classification. This study revealed promising results for CAD risk classification using machine learning models built on MPI SPECT radiomics. The proposed models are helpful to alleviate the labor-intensive MPI SPECT interpretation process regarding CAD status and can potentially expedite the diagnostic process.

Automatic Diagnosis of Stage of COVID-19 Patients using an Ensemble of Transfer Learning with Convolutional Neural Networks Based on Computed Tomography Images.

Background: Diagnosis of the stage of COVID-19 patients using the chest computed tomography (CT) can help the physician in making decisions on the length of time required for hospitalization and adequate selection of patient care. This diagnosis requires very expert radiologists who are not available everywhere and is also tedious and subjective. The aim of this study is to propose an advanced machine learning system to diagnose the stages of COVID-19 patients including normal, early, progressive, peak, and absorption stages based on lung CT images, using an automatic deep transfer learning ensemble. Methods: Different strategies of deep transfer learning were used which were based on pretrained convolutional neural networks (CNNs). Pretrained CNNs were fine-tuned on the chest CT images, and then, the extracted features were classified by a softmax layer. Finally, we built an ensemble method based on majority voting of the best deep transfer learning outputs to further improve the recognition performance. Results: The experimental results from 689 cases indicate that the ensemble of three deep transfer learning outputs based on EfficientNetB4, InceptionResV3, and NasNetlarge has the highest results in diagnosing the stage of COVID-19 with an accuracy of 91.66%. Conclusion: The proposed method can be used for the classification of the stage of COVID-19 disease with good accuracy to help the physician in making decisions on patient care.

Assessment of Field-in-Field, 3-Field, and 4-Field Treatment Planning Methods for Radiotherapy of Gastro-Esophageal Junction Cancer.

Background: Gastro-esophageal (GE) junction cancer is the fastest-growing tumor, particularly in the United States (US). Objective: This study aimed to compare dosimetric and radiobiological factors among field-in-field (FIF), three-field (3F), and four-field box (4FB) radiotherapy planning techniques for gastro-esophageal junction cancer. Material and Methods: In this experimental study, thirty patients with GE junction cancer were evaluated, and three planning techniques (field-in-field (FIF), three-field (3F), and four-field box (4FB)) were performed for each patient for a 6-MV photon beam. Dose distribution in the target volume, the monitor units (MUs) required, and the dose delivered to organs at risk (OARs) were compared for these techniques using the paired-sample t-test. Results: A significant difference was measured between the FIF and 3F techniques with respect to conformity index (CI), dose homogeneity index (HI), and tumor control probability (TCP) for the target organ, as well as the Dmean for the heart, kidneys, and liver. For the spinal cord, the FIF technique showed a slight reduction in the maximum dose compared to the other two techniques. In addition, the V20 Gy of the lungs and the normal tissue complication probability (NTCP) of all OARs were reduced with FIF method. Conclusion: The FIF technique showed better performance for treating patients with gastro-esophageal junction tumors, in terms of dose homogeneity in the target, conformity of the radiation field with the target volume, TCP, less dose to healthy organs, and fewer MU.

Dosimetry of small photon fields in the presence of bone heterogeneity using MAGIC polymer gel, Gafchromic film, and Monte Carlo simulation.

Background: The presence of heterogeneity within the radiation field increases the challenges of small field dosimetry. In this study, the performance of MAGIC polymer gel was evaluated in the dosimetry of small fields beyond bone heterogeneity. Materials and methods: Circular field sizes of 5, 10, 20 and 30 mm were used and Polytetrafluoroethylene with density of 2.2 g/cm3 was used as the bone equivalent material. The PDD curves, beam profiles, and penumbra widths were measured using MAGIC polymer gel, EBT2 film, and Monte Carlo simulation. Results: The maximum differences between MAGIC and EBT2 are 6.1, 4.7, 2.4, and 2.2 for PDD curves at 5, 10, 20, and 30 mm circular fields, respectively. The dose differences and distance to agreement between MAGIC and MC were within 1.89%/0.46 mm, 1.66%/0.43 mm, 1.28%/0.77 mm, and 1.31%/0.81 mm for beam profile values behind bone heterogeneity at 5, 10, 20, and 30 mm field sizes, respectively. Conclusion: The results presented that the MAGIC polymer gel dosimeter is a proper instrument for dosimetry beyond high density heterogeneity.

Detection of stage of lung changes in COVID-19 disease based on CT images: a radiomics approach.

The aim of this study is to classify patients suspected from COVID-19 to five stages as normal, early, progressive, peak, and absorption stages using radiomics approach based on lung computed tomography images. Lung CT scans of 683 people were evaluated. A set of statistical texture features was extracted from each CT image. The people were classified using the random forest algorithm as an ensemble method based on the decision trees outputs to five stages of COVID-19 disease. Proposed method attains the highest result with an accuracy of 93.55% (96.25% in normal, 74.39% in early, 100% in progressive, 82.19% in peak, and 96% in absorption stage) compared to the other three common classifiers. Radiomics method can be used for the classification of the stage of COVID-19 disease with good accuracy to help decide the length of time required to hospitalize patients, determine the type of treatment process required for patients in each category, and reduce the cost of care and treatment for hospitalized individuals.

Evaluation of ferrous benzoic methylthymol-blue gel as a dosimeter via magnetic resonance imaging.

PURPOSE: In previous studies, methylthymol-blue and benzoic acid have been introduced as a diffuser limiter and sensitivity enhancer in the gel dosimeter composition, respectively. This work focused on analyzing a formulation of the Fricke gel dosimeter consisting of methylthymol-blue and benzoic acid through magnetic resonance imaging. METHODS: The gel dosimeter samples were irradiated using 6, 10, and 15 MV photons with different levels of doses and read using a 1.5 T scanner in order to evaluate the dose-response sensitivity and to study the effect of benzoic acid concentration, diffusion coefficient and temperature and to determine the temporal stability of the gel dosimeter. RESULTS: Inspection of radiological properties revealed that this gel dosimeter can be considered as a tissue equivalent medium. Within the dose range 0 to 1000 cGy, the R1 sensitivity and R2 sensitivity of the gel dosimeter equaled 0.058 ± 0.003 and 0.092 ± 0.004 s-1Gy-1, respectively. The diffusion coefficient was less than 0.85 ± 0.02mm2h-1 for doses higher than 200 cGy. In addition, by changing the temperature from 15C to 25, the R1 sensitivity and R2 sensitivity decreased about 5 and 11%, respectively. Further, no significant energy and dose rate dependence were observed over photon energies of 6, 10, and 15 MV and over the range 65 to 525 cGy min-1. CONCLUSIONS: Based on our observation, the ferrous benzoic acid methylthymol-blue gel dosimeter can be suggested to measure the dose distribution. Further analysis is required to clarify its performance in clinical situations.

Gdn 3+@CNTs-PEG versus Gadovist®: In Vitro Assay.

OBJECTIVES: Carbon nanotubes (CNTs) are allotropes of carbon with a length-to-diameter ratio greater than 106 with the potential uses as medical diagnostic or therapeutic agents. In vitro studies have revealed that gadolinium (Gd) nanoparticle-catalyzed single-walled carbon nanotubes (SWCNTs) possess superparamagnetic properties, which enable them to be used as contrast agents in magnetic resonance imaging (MRI). Our study synthesized Gd-CNT for use as MRI contrast agents. METHODS: To reduce the toxicity and solubility of CNTs, it was functionalized, and after loading with Gd was coated with polyethylene glycols (PEG). We then synthesized different concentrations of Gdn 3+@CNTs-PEG and Gadovist® to be evaluated as MRI contrast agents. RESULTS: The analysis showed that the Gd concentration in Gadovist® was 12.18% higher than synthesized Gdn 3+@CNTs-PEG, but the mean signal intensity of the Gdn 3+@CNTs-PEG was approximately 3.3% times higher than Gadovist®. CONCLUSIONS: Our findings indicate that synthesized Gdn 3+@CNTs-PEG has the potential to be used as an MRI contrast agent in vitro, but in vivo assessment is necessary to determine the bio-distribution, kinetic, and signal enhancement characteristics.