Evaluating progress in automatic chest X-ray radiology report generation.

Artificial intelligence (AI) models for automatic generation of narrative radiology reports from images have the potential to enhance efficiency and reduce the workload of radiologists. However, evaluating the correctness of these reports requires metrics that can capture clinically pertinent differences. In this study, we investigate the alignment between automated metrics and radiologists' scoring of errors in report generation. We address the limitations of existing metrics by proposing new metrics, RadGraph F1 and RadCliQ, which demonstrate stronger correlation with radiologists' evaluations. In addition, we analyze the failure modes of the metrics to understand their limitations and provide guidance for metric selection and interpretation. This study establishes RadGraph F1 and RadCliQ as meaningful metrics for guiding future research in radiology report generation.

Estimation of white matter hyperintensities with synthetic MRI myelin volume fraction in patients with multiple sclerosis and non-multiple-sclerosis white matter hyperintensities: A pilot study among the Indian population.

AIM: Synthetic MRI (SyMRI) works on the MDME sequence, which acquires the relaxation properties of the brain and helps to measure the accurate tissue properties in 6 minutes. The aim of this study was to evaluate the synthetic MRI (SyMRI)-generated myelin (MyC) to white matter (WM) ratio, the WM fraction (WMF), MyC partial maps performing normative brain volumetry to investigate MyC loss in multiple sclerosis (MS) patients with white-matter hyperintensites (WMHs) and non-MS patients with WMHs in a clinical setting. MATERIALS and METHODS: Synthetic MRI images were acquired from 15 patients with MS, and from 15 non-MS patients on a 3T MRI scanner (Discovery MR750w; GE Healthcare; Milwaukee, USA) using MAGiC, a customized version of SyntheticMR's SyMRI® IMAGE software marketed by GE Healthcare under a license agreement. Fast multi-delay multi-echo acquisition was performed with a 2D axial pulse sequence with different combinations of echo time (TEs) and saturation delay times. The total image acquisition time was 6 minutes. SyMRI image analysis was done using SyMRI software (SyMRI Version: 11.3.6; Synthetic MR, Linköping, Sweden). SyMRI data were used to generate the MyC partial maps and WMFs to quantify the signal intensities of test group and control group, andcontrol group , and their mean values were recorded. All patients also underwent conventional diffusion-weighted imaging, i.e., T1w and T2w imaging. RESULTS: The results showed that the WMF was significantly lower in the test group than in the control group (38.8% vs 33.2%, p < 0.001). The Mann-Whitney U nonparametric t-test revealed a significant difference in the mean myelin volume between the test group and the control group (158.66 ± 32.31 vs. 138.29 ± 29.28, p = 0.044). Also, there were no significant differences in the gray matter fraction and intracranial volume between the test group and the control group. CONCLUSIONS: We observed MyC loss in test group using quantitative SyMRI. Thus, myelin loss in MS patients can be quantitatively evaluated using SyMRI.

Automated prostate multi-regional segmentation in magnetic resonance using fully convolutional neural networks.

OBJECTIVE: Automatic MR imaging segmentation of the prostate provides relevant clinical benefits for prostate cancer evaluation such as calculation of automated PSA density and other critical imaging biomarkers. Further, automated T2-weighted image segmentation of central-transition zone (CZ-TZ), peripheral zone (PZ), and seminal vesicle (SV) can help to evaluate clinically significant cancer following the PI-RADS v2.1 guidelines. Therefore, the main objective of this work was to develop a robust and reproducible CNN-based automatic prostate multi-regional segmentation model using an intercontinental cohort of prostate MRI. METHODS: A heterogeneous database of 243 T2-weighted prostate studies from 7 countries and 10 machines of 3 different vendors, with the CZ-TZ, PZ, and SV regions manually delineated by two experienced radiologists (ground truth), was used to train (n = 123) and test (n = 120) a U-Net-based model with deep supervision using a cyclical learning rate. The performance of the model was evaluated by means of dice similarity coefficient (DSC), among others. Segmentation results with a DSC above 0.7 were considered accurate. RESULTS: The proposed method obtained a DSC of 0.88 ± 0.01, 0.85 ± 0.02, 0.72 ± 0.02, and 0.72 ± 0.02 for the prostate gland, CZ-TZ, PZ, and SV respectively in the 120 studies of the test set when comparing the predicted segmentations with the ground truth. No statistically significant differences were found in the results obtained between manufacturers or continents. CONCLUSION: Prostate multi-regional T2-weighted MR images automatic segmentation can be accurately achieved by U-Net like CNN, generalizable in a highly variable clinical environment with different equipment, acquisition configurations, and population. KEY POINTS: • Deep learning techniques allows the accurate segmentation of the prostate in three different regions on MR T2w images. • Multi-centric database proved the generalization of the CNN model on different institutions across different continents. • CNN models can be used to aid on the diagnosis and follow-up of patients with prostate cancer.

A novel abnormality annotation database for COVID-19 affected frontal lung X-rays.

Consistent clinical observations of characteristic findings of COVID-19 pneumonia on chest X-rays have attracted the research community to strive to provide a fast and reliable method for screening suspected patients. Several machine learning algorithms have been proposed to find the abnormalities in the lungs using chest X-rays specific to COVID-19 pneumonia and distinguish them from other etiologies of pneumonia. However, despite the enormous magnitude of the pandemic, there are very few instances of public databases of COVID-19 pneumonia, and to the best of our knowledge, there is no database with annotation of abnormalities on the chest X-rays of COVID-19 affected patients. Annotated databases of X-rays can be of significant value in the design and development of algorithms for disease prediction. Further, explainability analysis for the performance of existing or new deep learning algorithms will be enhanced significantly with access to ground-truth abnormality annotations. The proposed COVID Abnormality Annotation for X-Rays (CAAXR) database is built upon the BIMCV-COVID19+ database which is a large-scale dataset containing COVID-19+ chest X-rays. The primary contribution of this study is the annotation of the abnormalities in over 1700 frontal chest X-rays. Further, we define protocols for semantic segmentation as well as classification for robust evaluation of algorithms. We provide benchmark results on the defined protocols using popular deep learning models such as DenseNet, ResNet, MobileNet, and VGG for classification, and UNet, SegNet, and Mask-RCNN for semantic segmentation. The classwise accuracy, sensitivity, and AUC-ROC scores are reported for the classification models, and the IoU and DICE scores are reported for the segmentation models.

A review on investigation of the basic contrast mechanism underlying multidimensional diffusion MRI in assessment of neurological disorders.

INTRODUCTION: Multidimensional diffusion MRI (MDD MRI) is a novel diffusion technique that uses advanced gradient waveforms for microstructural tissue characterization to provide information about average rate, anisotropy and orientation of the diffusion and to disentangle the signal fraction from specific cell types i.e., elongated cells, isotropic cells and free water. AIM: To review the diagnostic potential of MDD MRI in the clinical setting for microstructural tissue characterization in patients with neurological disorders to aid in patient care and treatment. METHOD: A scoping review on the clinical applications of MDD MRI was conducted from original articles published in PubMed and Scopus from 2015 to 2021 using the keywords "Multidimensional diffusion MRI" OR "diffusion tensor distribution" OR "Tensor-Valued Diffusion" OR "b-tensor encoding" OR "microscopic diffusion anisotropy" OR "microscopic anisotropy" OR "microscopic fractional anisotropy" OR "double diffusion encoding" OR "triple diffusion encoding" OR "double pulsed field gradients" OR "double wave vector" OR "correlation tensor imaging" AND "brain" OR "axons". RESULTS: Initially 145 articles were screened and after applying inclusion and exclusion criteria, nine articles were included in the final analysis. In most of these studies, microscopic diffusion anisotropy within the lesion showed deviation from the normal-appearing tissue. CONCLUSION: Multidimensional diffusion MRI can provide better quantification and visualization of tissue microstructure than conventional diffusion MRI and can be used in the clinical setting for diagnosis of neurological disorders.

Dapagliflozin Improves Body Fat Patterning, and Hepatic and Pancreatic Fat in Patients With Type 2 Diabetes in North India.

CONTEXT: Excess hepatic and pancreatic fat may contribute to hyperglycemia. OBJECTIVE: The objective of this study was to examine the effect of dapagliflozin (an SGLT2 inhibitor) on anthropometric profile, liver, and pancreatic fat in patients with type 2 diabetes mellitus (T2DM). METHODS: This is an observational interventional paired study design without a control group. Patients (n = 30) were given dapagliflozin 10 mg/day (on top of stable dose of metformin and/or sulfonylureas) for 120 days. Changes in anthropometry (circumferences and skinfold thickness), surrogate markers of insulin resistance, body composition, liver, and pancreatic fat (as measured by magnetic resonance imaging (MRI)-derived proton density fat fraction [FF]) were evaluated. RESULTS: After 120 days of treatment with dapagliflozin, a statistically significant reduction in weight, body mass index (BMI), body fat, circumferences, and all skinfold thickness was seen. A statistically significant reduction in blood glucose, glycated hemoglobin A1c, hepatic transaminases, fasting insulin, homeostatic model assessment of insulin resistance (HOMA-IR), and postprandial C-peptide was noted, while HOMA-ß, postprandial insulin sensitivity, and fasting adiponectin were statistically significantly increased. There was no change in lean body mass. Compared to baseline there was a statistically significant decrease in mean liver FF (from 15.2% to 10.1%, P < .0001) and mean pancreatic FF (from 7.5% to 5.99%, P < .0083). Reduction in liver fat was statistically significant after adjustment for change in body weight. CONCLUSION: Dapagliflozin, after 120 days of use, reduced pancreatic and liver fat and increased insulin sensitivity in Asian Indian patients with T2DM.

Clinical Explainability Failure (CEF) & Explainability Failure Ratio (EFR) - Changing the Way We Validate Classification Algorithms.

Adoption of Artificial Intelligence (AI) algorithms into the clinical realm will depend on their inherent trustworthiness, which is built not only by robust validation studies but is also deeply linked to the explainability and interpretability of the algorithms. Most validation studies for medical imaging AI report the performance of algorithms on study-level labels and lay little emphasis on measuring the accuracy of explanations generated by these algorithms in the form of heat maps or bounding boxes, especially in true positive cases. We propose a new metric - Explainability Failure Ratio (EFR) - derived from Clinical Explainability Failure (CEF) to address this gap in AI evaluation. We define an Explainability Failure as a case where the classification generated by an AI algorithm matches with study-level ground truth but the explanation output generated by the algorithm is inadequate to explain the algorithm's output. We measured EFR for two algorithms that automatically detect consolidation on chest X-rays to determine the applicability of the metric and observed a lower EFR for the model that had lower sensitivity for identifying consolidation on chest X-rays, implying that the trustworthiness of a model should be determined not only by routine statistical metrics but also by novel 'clinically-oriented' models.

Validation of expert system enhanced deep learning algorithm for automated screening for COVID-Pneumonia on chest X-rays.

SARS-CoV2 pandemic exposed the limitations of artificial intelligence based medical imaging systems. Earlier in the pandemic, the absence of sufficient training data prevented effective deep learning (DL) solutions for the diagnosis of COVID-19 based on X-Ray data. Here, addressing the lacunae in existing literature and algorithms with the paucity of initial training data; we describe CovBaseAI, an explainable tool using an ensemble of three DL models and an expert decision system (EDS) for COVID-Pneumonia diagnosis, trained entirely on pre-COVID-19 datasets. The performance and explainability of CovBaseAI was primarily validated on two independent datasets. Firstly, 1401 randomly selected CxR from an Indian quarantine center to assess effectiveness in excluding radiological COVID-Pneumonia requiring higher care. Second, curated dataset; 434 RT-PCR positive cases and 471 non-COVID/Normal historical scans, to assess performance in advanced medical settings. CovBaseAI had an accuracy of 87% with a negative predictive value of 98% in the quarantine-center data. However, sensitivity was 0.66-0.90 taking RT-PCR/radiologist opinion as ground truth. This work provides new insights on the usage of EDS with DL methods and the ability of algorithms to confidently predict COVID-Pneumonia while reinforcing the established learning; that benchmarking based on RT-PCR may not serve as reliable ground truth in radiological diagnosis. Such tools can pave the path for multi-modal high throughput detection of COVID-Pneumonia in screening and referral.

Model for in-vivo estimation of stiffness of tibiofemoral joint using MR imaging and FEM analysis.

BACKGROUND: Appropriate structural and material properties are essential for finite-element-modeling (FEM). In knee FEM, structural information could extract through 3D-imaging, but the individual subject's tissue material properties are inaccessible. PURPOSE: The current study's purpose was to develop a methodology to estimate the subject-specific stiffness of the tibiofemoral joint using finite-element-analysis (FEA) and MRI data of knee joint with and without load. METHODS: In this study, six Magnetic Resonance Imaging (MRI) datasets were acquired from 3 healthy volunteers with axially loaded and unloaded knee joint. The strain was computed from the tibiofemoral bone gap difference (ΔmBGFT) using the knee MR images with and without load. The knee FEM study was conducted using a subject-specific knee joint 3D-model and various soft-tissue stiffness values (1 to 50 MPa) to develop subject-specific stiffness versus strain models. RESULTS: Less than 1.02% absolute convergence error was observed during the simulation. Subject-specific combined stiffness of weight-bearing tibiofemoral soft-tissue was estimated with mean values as 2.40 ± 0.17 MPa. Intra-subject variability has been observed during the repeat scan in 3 subjects as 0.27, 0.12, and 0.15 MPa, respectively. All subject-specific stiffness-strain relationship data was fitted well with power function (R2 = 0.997). CONCLUSION: The current study proposed a generalized mathematical model and a methodology to estimate subject-specific stiffness of the tibiofemoral joint for FEM analysis. Such a method might enhance the efficacy of FEM in implant design optimization and biomechanics for subject-specific studies. Trial registration The institutional ethics committee (IEC), Indian Institute of Technology, Delhi, India, approved the study on 20th September 2017, with reference number P-019; it was a pilot study, no clinical trail registration was recommended.

The Algorithmic Audit: Working with Vendors to Validate Radiology-AI Algorithms-How We Do It.

There is a plethora of Artificial Intelligence (AI) tools that are being developed around the world aiming at either speeding up or improving the accuracy of radiologists. It is essential for radiologists to work with the developers of such algorithms to determine true clinical utility and risks associated with these algorithms. We present a framework, called an Algorithmic Audit, for working with the developers of such algorithms to test and improve the performance of the algorithms. The framework includes concepts of true independent validation on data that the algorithm has not seen before, curating datasets for such testing, deep examination of false positives and false negatives (to examine implications of such errors) and real-world deployment and testing of algorithms.

Unboxing AI - Radiological Insights Into a Deep Neural Network for Lung Nodule Characterization.

RATIONALE AND OBJECTIVES: To explain predictions of a deep residual convolutional network for characterization of lung nodule by analyzing heat maps. MATERIALS AND METHODS: A 20-layer deep residual CNN was trained on 1245 Chest CTs from National Lung Screening Trial (NLST) trial to predict the malignancy risk of a nodule. We used occlusion to systematically block regions of a nodule and map drops in malignancy risk score to generate clinical attribution heatmaps on 103 nodules from Lung Image Database Consortium image collection and Image Database Resource Initiative (LIDC-IDRI) dataset, which were analyzed by a thoracic radiologist. The features were described as heat inside nodule -bright areas inside nodule, peripheral heat continuous/interrupted bright areas along nodule contours, heat in adjacent plane -brightness in scan planes juxtaposed with the nodule, satellite heat - a smaller bright spot in proximity to nodule in the same scan plane, heat map larger than nodule bright areas corresponding to the shape of the nodule seen outside the nodule margins and heat in calcification. RESULTS: These six features were assigned binary values. This feature vector was fedinto a standard J48 decision tree with 10-fold cross-validation, which gave an 85 % weighted classification accuracy with a 77.8% True Positive (TP) rate, 8% False Positive (FP) rate for benign cases and 91.8% TP and 22.2% FP rates for malignant cases. Heat Inside nodule was more frequently observed in nodules classified as malignant whereas peripheral heat, heat in adjacent plane, and satellite heat were more commonly seen in nodules classified as benign. CONCLUSION: We discuss the potential ability of a radiologist to visually parse the deep learning algorithm generated "heat map" to identify features aiding classification.

Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study.

BACKGROUND: Non-contrast head CT scan is the current standard for initial imaging of patients with head trauma or stroke symptoms. We aimed to develop and validate a set of deep learning algorithms for automated detection of the following key findings from these scans: intracranial haemorrhage and its types (ie, intraparenchymal, intraventricular, subdural, extradural, and subarachnoid); calvarial fractures; midline shift; and mass effect. METHODS: We retrospectively collected a dataset containing 313 318 head CT scans together with their clinical reports from around 20 centres in India between Jan 1, 2011, and June 1, 2017. A randomly selected part of this dataset (Qure25k dataset) was used for validation and the rest was used to develop algorithms. An additional validation dataset (CQ500 dataset) was collected in two batches from centres that were different from those used for the development and Qure25k datasets. We excluded postoperative scans and scans of patients younger than 7 years. The original clinical radiology report and consensus of three independent radiologists were considered as gold standard for the Qure25k and CQ500 datasets, respectively. Areas under the receiver operating characteristic curves (AUCs) were primarily used to assess the algorithms. FINDINGS: The Qure25k dataset contained 21 095 scans (mean age 43 years; 9030 [43%] female patients), and the CQ500 dataset consisted of 214 scans in the first batch (mean age 43 years; 94 [44%] female patients) and 277 scans in the second batch (mean age 52 years; 84 [30%] female patients). On the Qure25k dataset, the algorithms achieved an AUC of 0·92 (95% CI 0·91-0·93) for detecting intracranial haemorrhage (0·90 [0·89-0·91] for intraparenchymal, 0·96 [0·94-0·97] for intraventricular, 0·92 [0·90-0·93] for subdural, 0·93 [0·91-0·95] for extradural, and 0·90 [0·89-0·92] for subarachnoid). On the CQ500 dataset, AUC was 0·94 (0·92-0·97) for intracranial haemorrhage (0·95 [0·93-0·98], 0·93 [0·87-1·00], 0·95 [0·91-0·99], 0·97 [0·91-1·00], and 0·96 [0·92-0·99], respectively). AUCs on the Qure25k dataset were 0·92 (0·91-0·94) for calvarial fractures, 0·93 (0·91-0·94) for midline shift, and 0·86 (0·85-0·87) for mass effect, while AUCs on the CQ500 dataset were 0·96 (0·92-1·00), 0·97 (0·94-1·00), and 0·92 (0·89-0·95), respectively. INTERPRETATION: Our results show that deep learning algorithms can accurately identify head CT scan abnormalities requiring urgent attention, opening up the possibility to use these algorithms to automate the triage process. FUNDING: Qure.ai.

Low-Grade Central Osteosarcoma: Report of Two Unusual Morphologic Variants.

Low-grade central osteosarcoma (LGCO) is a rare subtype of osteosarcoma, constituting < 2% of all osteosarcomas. If not treated appropriately, the tumor can recur with higher-grade disease. We report two cases of low-grade central osteosarcoma with unusual morphologic features and belonging to different age groups. Both presented with pain and swelling in the lower end of femur. Radiologically, both the lesions revealed a large mass with irregular borders and soft tissue invasion. One patient underwent above-knee amputation and wide local excision of tumor was done in the other patient. Histologically, both the tumors showed spindle cell proliferation displaying mild atypia. In synopsis with radiology, diagnosis of low-grade central osteosarcoma was made in both cases. These cases highlight the varied morphological spectrum of low-grade central osteosarcoma and underscore the diagnostic difficulties faced. Recognition of the variants of low-grade central osteosarcoma is based on aggressive radiological appearance and on adequate tumor sampling for histologic examination.

Hydatid cyst of parotid: Report of unusual cytological findings extending the cytomorphological spectrum.

Hydatid disease is a zoonotic infestation caused by larval cestode of genus Echinococcus. Cystic form of this infection mostly involves liver and lung. Hydatid disease of the parotid gland is very rare even in endemic areas and is often clinically mistaken for parotid tumors or cysts. The presence of protoscolices, laminated membranes, and isolated hooklets are characteristic cytological features observed on fine-needle aspirate from hydatid cysts. We report unusual cytological features from a hydatid cyst of parotid in a 13-year-old girl. She presented with a slowly enlarging hard mass in left parotid. Fine-needle aspiration yielded slightly turbid fluid. Smears from the sediment revealed naked parasitic micronuclei, fragments of germinative layer (endocyst), and abortive brood capsules (buds from endocyst) seen as spherical structures with multiple parasitic nuclei. Some of these spherical structures were degenerated with recognizable nuclei and others were completely necrotic. Diagn. Cytopathol. 2016;44:770-773. © 2016 Wiley Periodicals, Inc.