Artificial intelligence-based analysis of whole-body bone scintigraphy: The quest for the optimal deep learning algorithm and comparison with human observer performance.

PURPOSE: Whole-body bone scintigraphy (WBS) is one of the most widely used modalities in diagnosing malignant bone diseases during the early stages. However, the procedure is time-consuming and requires vigour and experience. Moreover, interpretation of WBS scans in the early stages of the disorders might be challenging because the patterns often reflect normal appearance that is prone to subjective interpretation. To simplify the gruelling, subjective, and prone-to-error task of interpreting WBS scans, we developed deep learning (DL) models to automate two major analyses, namely (i) classification of scans into normal and abnormal and (ii) discrimination between malignant and non-neoplastic bone diseases, and compared their performance with human observers. MATERIALS AND METHODS: After applying our exclusion criteria on 7188 patients from three different centers, 3772 and 2248 patients were enrolled for the first and second analyses, respectively. Data were split into two parts, including training and testing, while a fraction of training data were considered for validation. Ten different CNN models were applied to single- and dual-view input (posterior and anterior views) modes to find the optimal model for each analysis. In addition, three different methods, including squeeze-and-excitation (SE), spatial pyramid pooling (SPP), and attention-augmented (AA), were used to aggregate the features for dual-view input models. Model performance was reported through area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, and specificity and was compared with the DeLong test applied to ROC curves. The test dataset was evaluated by three nuclear medicine physicians (NMPs) with different levels of experience to compare the performance of AI and human observers. RESULTS: DenseNet121_AA (DensNet121, with dual-view input aggregated by AA) and InceptionResNetV2_SPP achieved the highest performance (AUC = 0.72) for the first and second analyses, respectively. Moreover, on average, in the first analysis, Inception V3 and InceptionResNetV2 CNN models and dual-view input with AA aggregating method had superior performance. In addition, in the second analysis, DenseNet121 and InceptionResNetV2 as CNN methods and dual-view input with AA aggregating method achieved the best results. Conversely, the performance of AI models was significantly higher than human observers for the first analysis, whereas their performance was comparable in the second analysis, although the AI model assessed the scans in a drastically lower time. CONCLUSION: Using the models designed in this study, a positive step can be taken toward improving and optimizing WBS interpretation. By training DL models with larger and more diverse cohorts, AI could potentially be used to assist physicians in the assessment of WBS images.

Diagnostic performance of prone-only myocardial perfusion imaging versus coronary angiography in the detection of coronary artery disease: A systematic review and meta-analysis.

STUDY DESIGN: Although prone position is considered as a complementary protocol in myocardial perfusion imaging (MPI), there is no consensus on its capability to find coronary artery disease (CAD), independently. The primary aim of this review was to report pooled sensitivity and specificity for prone position MPI in detection of CAD. In addition, the results were compared to the supine position's performance. METHODS: Electronic bibliographic databases, The Cochrane Library, Web of Science (Science and Social Science Citation Index), Scopus, PubMed, and EMBASE until the end of June 2020 were searched. Studies were included based on the inclusion criteria of (1) evaluated the prone position MPI, (2) defined CAD with coronary angiography (CAG), using the threshold of ≥ 50% stenosis, (3) Adequate data were provided to extract the diagnostic performance. QUADAS-2 tool was utilized to assess the quality of included studies. Pooled sensitivity and specificity were calculated for prone and supine positions, separately. The hierarchical summary ROC curves were also drawn. RESULTS: Ten individual studies with the data of the 1490 patients for the prone position and 1138 patients for the supine position were included. Pooled sensitivity and specificity for the prone position were 83% and 79%, respectively. These results were calculated for the supine position as the sensitivity of 86% and specificity of 67%. The pooled sensitivity and specificity of the prone position in detecting the right coronary artery territory defects were 70% and 84%, in turn. CONCLUSION: In the suspicion for the CAD, prone position with comparable sensitivity and higher specificity can be an acceptable alternative to the supine position as the standard method. Also, in the cases of possible defects in the RCA territory, prone position showed to be a superior standard.

Role of Technetium-99m-Ethylenedicysteine in the Calculation of Differential Renal Function: A Comparison Study with Dimercaptosuccinic Acid Renal Scintigraphy.

PURPOSE OF THE STUDY: The aim of our study was to compare the technetium-99m (Tc-99m)-ethylenedicysteine (EC) renography calculation of differential renal function (DRF) with this measurement using Tc-99m-dimercaptosuccinic acid (DMSA) scintigraphy. MATERIALS AND METHODS: Patients referred to our department were included in our study, and both DMSA and EC scans were performed for each patient according to the standard imaging protocols. A checklist was filled for each patient. Statistical analysis was performed using correlation and regression methods. RESULTS: Forty-two patients (mean age: 3.6 ± 3.4 years), including 32 boys and 10 girls, participated in our study. The results of EC scintigraphy were significantly correlated with the values of DMSA scintigraphy (P < 0.001). Performing linear regression, EC renography significantly (P < 0.001) predicted the DRF as it was calculated by DMSA scintigraphy (R 2 = 0.92, P < 0.001). This test was significant in both male and female subgroups (P < 0.001). CONCLUSIONS: Overall, our study findings were similar to the reported results in the other reviewed studies, showing that Tc-99m-EC can be considered as an alternative for DMSA scintigraphy, providing interchangeable results.

Determination of the value of albumin, anti-thrombin III, fibrinogen and D-dimer factors in the diagnosis of asymptomatic pulmonary embolism in patients with nephrotic syndrome.

BACKGROUND: Thromboembolism is one of the most important and dangerous complications of nephrotic syndrome. This study aimed to determine the value of albumin, anti-thrombin III, fibrinogen and D-dimer factors in the prediction of asymptomatic pulmonary embolism in patients with nephrotic syndrome in non-remission period. METHODS: Plasma levels of albumin, anti-thrombin III, fibrinogen and D-dimer were assessed in 30 nephrotic children in non-remission period (including new case-patient or relapse period), and the results were compared with chest X-ray and lung perfusion scintigraphy (Q scan). RESULTS: The mean age of patients was 6.22 ± 3.5 years (range 2-12 years). Of patients, 23.3% had abnormal findings in perfusion scan suggestive of pulmonary emboli despite absence of any respiratory manifestations. Median plasma albumin and anti-thrombin III levels in patients with asymptomatic pulmonary embolism were lower than in patients without pulmonary embolism. Also, median fibrinogen and D-dimer levels in patients with asymptomatic pulmonary embolism were higher than in patients without pulmonary embolism, with no statistically significant differences between sex, age, hemoglobin and hematocrit of patients and lung perfusion scan results. CONCLUSION: Patients with abnormal blood levels of albumin (< 3.5 g/dl), anti-thrombin III (< 80 ml/dl), fibrinogen (> 400 ml/dl) and D-dimer (> 0.5 µg/dl) underwent CXR/Q scan and were treated with heparin if there was pulmonary embolism.

CAD system based on B-mode and color Doppler sonographic features may predict if a thyroid nodule is hot or cold.

OBJECTIVES: The aim of this study was to evaluate if the analysis of sonographic parameters could predict if a thyroid nodule was hot or cold. METHODS: Overall, 102 thyroid nodules, including 51 hyperfunctioning (hot) and 51 hypofunctioning (cold) nodules, were evaluated in this study. Twelve sonographic features (i.e., seven B-mode and five Doppler features) were extracted for each nodule type. The isthmus thickness, nodule volume, echogenicity, margin, internal component, microcalcification, and halo sign features were obtained in the B-mode, while the vascularity pattern, resistive index (RI), peak systolic velocity, end diastolic velocity, and peak systolic/end diastolic velocity ratio (SDR) were determined, based on Doppler ultrasounds. All significant features were incorporated in the computer-aided diagnosis (CAD) system to classify hot and cold nodules. RESULTS: Among all sonographic features, only isthmus thickness, nodule volume, echogenicity, RI, and SDR were significantly different between hot and cold nodules. Based on these features in the training dataset, the CAD system could classify hot and cold nodules with an area under the curve (AUC) of 0.898. Also, in the test dataset, hot and cold nodules were classified with an AUC of 0.833. CONCLUSIONS: 2D sonographic features could differentiate hot and cold thyroid nodules. The CAD system showed a great potential to achieve it automatically. KEY POINTS: • Cold nodules represent higher volume (p = 0.005), isthmus thickness (p = 0.035), RI (p = 0.020), and SDR (p = 0.044) and appear hypoechogenic (p = 0.010) in US. • Nodule volume with an AUC of 0.685 and resistive index with an AUC of 0.628 showed the highest classification potential among all B-mode and Doppler features respectively. • The proposed CAD system could distinguish hot nodules from cold ones with an AUC of 0.833 (sensitivity 90.00%, specificity 70.00%, accuracy 80.00%, PPV 87.50%, and NPV 75.00%).

Scintigraphic texture analysis for assessment of renal allograft function.

PURPOSE: Early detection and monitoring of kidney function during the post-transplantation period is one of the most important issues for improving the accuracy of an initial diagnosis. The aim of this study was to evaluate texture analysis (TA) in scintigraphic imaging to detect changes in kidney status after transplantation. MATERIAL AND METHODS: Scintigraphic images were used for TA from a total of 94 kidney allografts (39 rejected and 55 non-rejected). Images corresponding to the frames at the 2nd, 5th, and 20th minute of the study were used to determine the optimum time point for analysis of differences in texture features between the rejected and non-rejected allografts. RESULTS: Linear discriminant analysis indicated the best performance at the fifth minute frame for classification of the rejected and non-rejected allografts with receiver operating characteristic curve (Az) of 0.982, corresponding to 91.89% sensitivity, 96.49% specificity, and 94.68% accuracy. Also, TA can differentiate acute tubular necrosis from acute rejection with Az of 0.953 corresponding to 88% sensitivity, 92.31% specificity, and 90.62% accuracy at the 5th minute frame. The best correlation between texture feature and kidney function was achieved at the 20th minute frame (r = -0.396) for glomerular filtration rate. CONCLUSIONS: TA has good potential for the characterisation of kidney failure after transplantation and can improve clinical diagnosis.

COMPARE CPM-RMI Trial: Intramyocardial Transplantation of Autologous Bone Marrow-Derived CD133+ Cells and MNCs during CABG in Patients with Recent MI: A Phase II/III, Multicenter, Placebo-Controlled, Randomized, Double-Blind Clinical Trial.

This article published in Cell J (Yakhteh), Vol 20, No 2, Jul-Sep 2018, on pages 267-277, four affiliations (1, 4, 5, and 10) were changed based on authors request.

COMPARE CPM-RMI Trial: Intramyocardial Transplantation of Autologous Bone Marrow-Derived CD133+ Cells and MNCs during CABG in Patients with Recent MI: A Phase II/III, Multicenter, Placebo-Controlled, Randomized, Double-Blind Clinical Trial.

OBJECTIVES: The regenerative potential of bone marrow-derived mononuclear cells (MNCs) and CD133+ stem cells in the heart varies in terms of their pro-angiogenic effects. This phase II/III, multicenter and double-blind trial is designed to compare the functional effects of intramyocardial autologous transplantation of both cell types and placebo in patients with recent myocardial infarction (RMI) post-coronary artery bypass graft. MATERIALS AND METHODS: This was a phase II/III, randomized, double-blind, placebo-controlled trial COMPARE CPM-RMI (CD133, Placebo, MNCs - recent myocardial infarction) conducted in accordance with the Declaration of Helsinki that assessed the safety and efficacy of CD133 and MNCs compared to placebo in patients with RMI. We randomly assigned 77 eligible RMI patients selected from 5 hospitals to receive CD133+ cells, MNC, or a placebo. Patients underwent gated single photon emission computed tomography assessments at 6 and 18 months post-intramyocardial transplantation. We tested the normally distributed efficacy outcomes with a mixed analysis of variance model that used the entire data set of baseline and between-group comparisons as well as within subject (time) and group×time interaction terms. RESULTS: There were no related serious adverse events reported. The intramyocardial transplantation of both cell types increased left ventricular ejection fraction by 9% [95% confidence intervals (CI): 2.14% to 15.78%, P=0.01] and improved decreased systolic wall thickening by -3.7 (95% CI: -7.07 to -0.42, P=0.03). The CD133 group showed significantly decreased non-viable segments by 75% (P=0.001) compared to the placebo and 60% (P=0.01) compared to the MNC group. We observed this improvement at both the 6- and 18-month time points. CONCLUSIONS: Intramyocardial injections of CD133+ cells or MNCs appeared to be safe and efficient with superiority of CD133+ cells for patients with RMI. Although the sample size precluded a definitive statement about clinical outcomes, these results have provided the basis for larger studies to confirm definitive evidence about the efficacy of these cell types (Registration Number: NCT01167751).

Predictive quantitative sonographic features on classification of hot and cold thyroid nodules.

PURPOSE: This study investigated the potentiality of ultrasound imaging to classify hot and cold thyroid nodules on the basis of textural and morphological analysis. METHODS: In this research, 42 hypo (hot) and 42 hyper-function (cold) thyroid nodules were evaluated through the proposed method of computer aided diagnosis (CAD) system. To discover the difference between hot and cold nodules, 49 sonographic features (9 morphological, 40 textural) were extracted. A support vector machine classifier was utilized for the classification of LNs based on their extracted features. RESULTS: In the training set data, a combination of morphological and textural features represented the best performance with area under the receiver operating characteristic curve (AUC) of 0.992. Upon testing the data set, the proposed model could classify the hot and cold thyroid nodules with an AUC of 0.948. CONCLUSIONS: CAD method based on textural and morphological features is capable of distinguishing between hot from cold nodules via 2-Dimensional sonography. Therefore, it can be used as a supplementary technique in daily clinical practices to improve the radiologists' understanding of conventional ultrasound imaging for nodules characterization.

Fragmented QRS on surface electrocardiogram as a predictor of perfusion defect in patients with suspected coronary artery disease undergoing myocardial perfusion imaging.

BACKGROUND: Fragmented QRS (FQRS) on surface electrocardiogram (ECG) is associated with the presence of myocardial scar tissue and may have prognostic value after certain ischemic events. We aimed to examine the anatomical correlation of FQRS with the presence of perfusion abnormalities in patients with suspected coronary artery disease (CAD). METHODS: Patients without a known history of CAD, who were referred for myocardial perfusion imaging (MPI) between January 2016 and May 2016, were enrolled. The presence of FQRS on surface ECG was evaluated. The presence of FQRS, number of leads with FQRS, and the location of FQRS as well as patient characteristics were compared in patients with normal versus abnormal MPI. Multivariate model was constructed to identify independent factors associated with perfusion defect. RESULTS: One hundred four women and 94 men were enrolled. Fragmentation of anterior, lateral, and inferior leads was detected in 13 (6.5%), 17 (8.5%), and 36 (18.1%) subjects, respectively. MPI was normal in 134 (67.6%) patients. FQRS was significantly more common in patients with abnormal MPI (p < 0.001). Age (odds ratio [OR]: 1.05 [1.02-1.08]; p = 0.001), number of the leads presenting FQRS (OR: 1.46 [1.12-1.92] p = 0.006), and diabetes (OR: 2.33 [1.16-4.69]; p = 0.018) were independent predictors of the presence of perfusion defect on MPI. CONCLUSION: In the absence of known CAD, FQRS is associated with the presence of perfusion abnormalities. Incorporating FQRS in diagnostic armamentarium may aid us in selecting patients who may benefit from MPI.

Solitary metastasis of gastric cancer to fibula: a case report.

Gastric cancer is one of the most common and most fatal neoplasms in human. Its skeletal metastasis is less frequent, particularly when solitary. The objective of this article is to represent a case of solitary fibular metastasis from this cancer not reported before based on Medline search.

Urine macrophage migration inhibitory factor (MIF) in children with urinary tract infection: a possible predictor of acute pyelonephritis.

Macrophage migration inhibitory factor (MIF) is an important pro-inflammatory cytokine expressed at sites of inflammation. We have assessed this factor (MIF) in urinary tract infections with the aim of determining a non-invasive and sensitive method to differentiate upper and lower renal involvement. Thirty-three pediatric patients with urinary track infection (25 with acute pyelonephritis, eight with acute cystitis) and 40 healthy subjects were recruited for this prospective case-control study. Pyelonephritis was differentiated from cystitis by dimercaptosuccinic acid (DMSA) scan. Urinary MIF concentration was determined using an enzyme-linked immunosorbent assay method. The urine MIF/creatinine (Cr) ratio was significantly higher in pyelonephritis patients than in those with acute cystitis and the control group (P < 0.001). The optimal cut-point of 4.90 pg/micromol Cr for the urine MIF/Cr ratio has the potential to be a biomarker for distinguishing patients with acute pyelonephritis from those with acute cystitis. Determination of the urinary MIF was also useful in selecting the patients at risk of permanent renal damage. Of those patients with pyelonephritis, based on the DMSA scan at the time of infection, scarring on follow-up DMSA scan 9-12 months later occurred in patients with the highest urinary MIF/Cr ratios. We conclude that the urine MIF/Cr ratio is a sensitive test for differentiating acute pyelonephritis from acute cystitis and also for detecting children with acute pyelonephritis who are at a higher risk for permanent renal scars in the future.

Ethylenedicysteine versus diethylenetriamine pentaacetic acid as the carrier of technetium Tc 99m in diuretic renography for patients with upper urinary tract obstruction.

INTRODUCTION: L,L-ethylenedicysteine (EC) is a new carrier of technetium Tc 99m (99mTc) with a lower affinity to plasma albumin in comparison with diethylenetriamine pentaacetic acid (DTPA). We compared 99mTc-EC scan with 99mTc-DTPA scan in diuretic renography for patients with obstructive uropathy. MATERIALS AND METHODS: Thirty-three patients with upper urinary tract obstruction were randomly selected and underwent diuretic renographies by 99mTc-EC and 99mTc-DTPA. The counts of radioisotope per pixel in the target (the kidney) and background tissues as well as the clearance half-life of these two radiopharmaceuticals were measured and compared. RESULTS: Mean counts of radioisotope per pixel in the target tissue was not different between 99mTc-EC and 99mTc-DTPA scans, but in the background tissue, it was less for 99mTc-EC (P = .003). Target-background ratio was higher for 99mTc-EC scan (3.80 +/- 2.11 versus 2.48 +/- 1.39; P < .001). Renal clearance half-life of radioisotope was shorter for 99mTc-EC scan than 99mTc-DTPA scan (58.15 +/- 15.17 minutes versus 78.65 +/- 19.99 minutes; P = .033). The results were similar for uremic patients (with a serum creatinine level > 2mg/dL). CONCLUSION: Target-background ratio of radiopharmaceutical uptake rates in diuretic renography was a good indicator of the higher resolution of 99mTc-EC than 99mTc-DTPA scan. We also demonstrated the faster clearance of 99mTc-EC than 99mTc-DTPA. This results in less radiation that is especially useful in children. To our opinion, 99mTc-EC can better depict the kidneys in comparison with 99mTc-DTPA.