Artificial intelligence (AI) is gaining increasing importance in many medical specialties, yet data on patients' opinions on the use of AI in medicine are scarce.
This study aimed to investigate ...patients' opinions on the use of AI in different aspects of the medical workflow and the level of control and supervision under which they would deem the application of AI in medicine acceptable.
Patients scheduled for computed tomography or magnetic resonance imaging voluntarily participated in an anonymized questionnaire between February 10, 2020, and May 24, 2020. Patient information, confidence in physicians vs AI in different clinical tasks, opinions on the control of AI, preference in cases of disagreement between AI and physicians, and acceptance of the use of AI for diagnosing and treating diseases of different severity were recorded.
In total, 229 patients participated. Patients favored physicians over AI for all clinical tasks except for treatment planning based on current scientific evidence. In case of disagreement between physicians and AI regarding diagnosis and treatment planning, most patients preferred the physician's opinion to AI (96.2% 153/159 vs 3.8% 6/159 and 94.8% 146/154 vs 5.2% 8/154, respectively; P=.001). AI supervised by a physician was considered more acceptable than AI without physician supervision at diagnosis (confidence rating 3.90 SD 1.20 vs 1.64 SD 1.03, respectively; P=.001) and therapy (3.77 SD 1.18 vs 1.57 SD 0.96, respectively; P=.001).
Patients favored physicians over AI in most clinical tasks and strongly preferred an application of AI with physician supervision. However, patients acknowledged that AI could help physicians integrate the most recent scientific evidence into medical care. Application of AI in medicine should be disclosed and controlled to protect patient interests and meet ethical standards.
Deep learning-based algorithms have demonstrated enormous performance in segmentation of medical images. We collected a dataset of multiparametric MRI and contour data acquired for use in ...radiosurgery, to evaluate the performance of deep convolutional neural networks (DCNN) in automatic segmentation of brain metastases (BM).
A conventional U-Net (cU-Net), a modified U-Net (moU-Net) and a U-Net trained only on BM smaller than 0.4 ml (sU-Net) were implemented. Performance was assessed on a separate test set employing sensitivity, specificity, average false positive rate (AFPR), the dice similarity coefficient (DSC), Bland-Altman analysis and the concordance correlation coefficient (CCC).
A dataset of 509 patients (1223 BM) was split into a training set (469 pts) and a test set (40 pts). A combination of all trained networks was the most sensitive (0.82) while maintaining a specificity 0.83. The same model achieved a sensitivity of 0.97 and a specificity of 0.94 when considering only lesions larger than 0.06 ml (75% of all lesions). Type of primary cancer had no significant influence on the mean DSC per lesion (p = 0.60). Agreement between manually and automatically assessed tumor volumes as quantified by a CCC of 0.87 (95% CI, 0.77-0.93), was excellent.
Using a dataset which properly captured the variation in imaging appearance observed in clinical practice, we were able to conclude that DCNNs reach clinically relevant performance for most lesions. Clinical applicability is currently limited by the size of the target lesion. Further studies should address if small targets are accurately represented in the test data.
Objectives
To evaluate the correlation between simple planimetric measurements in axial computed tomography (CT) slices and measurements of patient body composition and anthropometric data performed ...with bioelectrical impedance analysis (BIA) and metric clinical assessments.
Methods
In this prospective cross-sectional study, we analyzed data of a cohort of 62 consecutive, untreated adult patients with advanced malignant melanoma who underwent concurrent BIA assessments at their radiologic baseline staging by CT between July 2016 and October 2017. To assess muscle and adipose tissue mass, we analyzed the areas of the paraspinal muscles as well as the cross-sectional total patient area in a single CT slice at the height of the third lumbar vertebra. These measurements were subsequently correlated with anthropometric (body weight) and body composition parameters derived from BIA (muscle mass, fat mass, fat-free mass, and visceral fat mass). Linear regression models were built to allow for estimation of each parameter based on CT measurements.
Results
Linear regression models allowed for accurate prediction of patient body weight (adjusted
R
2
= 0.886), absolute muscle mass (adjusted
R
2
= 0.866), fat-free mass (adjusted
R
2
= 0.855), and total as well as visceral fat mass (adjusted
R
2
= 0.887 and 0.839, respectively).
Conclusions
Our data suggest that patient body composition can accurately and quantitatively be determined by using simple measurements in a single axial CT slice. This could be useful in various medical and scientific settings, where the knowledge of the patient’s anthropometric parameters is not immediately or easily available.
Key Points
•
Easy to perform measurements on a single CT slice highly correlate with clinically valuable parameters of body composition.
•
Body composition data were acquired using bioelectrical impedance analysis to correlate CT measurements with a non-imaging-based method, which is frequently lacking in previous studies.
•
The obtained equations facilitate a quick, opportunistic assessment of relevant parameters of body composition.
Meningioma grading is relevant to therapy decisions in complete or partial resection, observation, and radiotherapy because higher grades are associated with tumor growth and recurrence. The ...differentiation of low and intermediate grades is particularly challenging. This study attempts to apply radiomics-based shape and texture analysis on routine multiparametric magnetic resonance imaging (MRI) from different scanners and institutions for grading.
We used MRI data (T1-weighted/T2-weighted, T1-weighted-contrast-enhanced T1CE, fluid-attenuated inversion recovery FLAIR, diffusion-weighted imaging DWI, apparent diffusion coefficient ADC) of grade I (n = 46) and grade II (n = 25) nontreated meningiomas with histologic workup. Two experienced radiologists performed manual tumor segmentations on FLAIR, T1CE, and ADC images in consensus. The MRI data were preprocessed through T1CE and T1-subtraction, coregistration, resampling, and normalization. A PyRadiomics package was used to generate 990 shape/texture features. Stepwise dimension reduction and robust radiomics feature selection were performed. Biopsy results were used as standard of reference.
Four statistically independent radiomics features were identified as showing the strongest predictive values for higher tumor grades: roundness-of-FLAIR-shape (area under curve AUC, 0.80), cluster-shades-of-FLAIR/T1CE-gray-level (AUC, 0.80), DWI/ADC-gray-level-variability (AUC, 0.72), and FLAIR/T1CE-gray-level-energy (AUC, 0.76). In a multivariate logistic regression model, the combination of the features led to an AUC of 0.91 for the differentiation of grade I and grade II meningiomas.
Our results indicate that radiomics-based feature analysis applied on routine MRI is viable for meningioma grading, and a multivariate logistic regression model yielded strong classification performances. More advanced tumor stages are identifiable through certain shape parameters of the lesion, textural patterns in morphologic MRI sequences, and DWI/ADC variability.
Objectives
This study compares reduction of strong metal artifacts from large dental implants/bridges using spectral detector CT-derived virtual monoenergetic images (VMI), metal artifact reduction ...algorithms/reconstructions (MAR), and a combination of both methods (VMI
MAR
) to conventional CT images (CI).
Methods
Forty-one spectral detector CT (SDCT) datasets of patients that obtained additional MAR reconstructions due to strongest artifacts from large oral implants were included. CI, VMI, MAR, and VMI
MAR
ranging from 70 to 200 keV (10 keV increment) were reconstructed. Objective image analyses were performed ROI-based by measurement of attenuation (HU) and standard deviation in most pronounced hypo-/hyperdense artifacts as well as artifact impaired soft tissue (mouth floor/soft palate). Extent of artifact reduction, diagnostic assessment of soft tissue, and appearance of new artifacts were rated visually by two radiologists.
Results
The hypo-/hyperattenuating artifacts showed an increase and decrease of HU values in MAR and VMI
MAR
(CI/MAR/VMI
MAR-200keV
: − 369.8 ± 239.6/− 37.3 ± 109.6/− 46.2 ± 71.0 HU,
p
< 0.001 and 274.8 ± 170.2/51.3 ± 150.8/36.6 ± 56.0,
p
< 0.001, respectively). Higher keV values in hyperdense artifacts allowed for additional artifact reduction; however, this trend was not significant. Artifacts in soft tissue were reduced significantly by MAR and VMI
MAR
. Visually, high-keV VMI, MAR, and VMI
MAR
reduced artifacts and improved diagnostic assessment of soft tissue. Overcorrection/new artifacts were reported that mostly did not hamper diagnostic assessment. Overall interrater agreement was excellent (ICC = 0.85).
Conclusions
In the presence of strong artifacts due to large oral implants, MAR is a powerful mean for artifact reduction. For hyperdense artifacts, MAR should be supplemented by VMI ranging from 140 to 200 keV. This combination yields optimal artifact reduction and improves the diagnostic image assessment in imaging of the head and neck.
Key Points
• Large oral implants can cause strong artifacts.
• MAR is a powerful tool for artifact reduction considering such strong artifacts.
• Hyperdense artifact reduction is supplemented by VMI of 140–200 keV from SDCT.
Highlights • Mono-energetic reconstructions in dual-layer spectral detector CT improve image quality compared to poly-energetic images. • Noise levels and CNR were better for mono-energetic ...reconstructions compared to conventional images in contrast-enhanced chest CT. • Whether mono-energetic images have added diagnostic value compared to conventional images, is still unknown.
OBJECTIVESDual-energy computed tomography (DECT)–derived quantification of iodine concentration (IC) is increasingly used in oncologic imaging to characterize lesions and evaluate treatment response. ...However, only limited data are available on intraindividual consistency of IC and its variation. This study investigates the longitudinal reproducibility of IC in organs, vessels, and lymph nodes in a large cohort of healthy patients who underwent repetitive DECT imaging.
MATERIALS AND METHODSA total of 159 patients, who underwent a total of 469 repetitive (range, 2–4), clinically indicated portal-venous phase DECT examinations of the chest and abdomen, were retrospectively included. At time of imaging, macroscopic tumor burden was excluded by follow-up imaging (≥3 months). Iodine concentration was measured region of interest-based (N = 43) in parenchymatous organs, vessels, lymph nodes, and connective tissue. Normalization of IC to the aorta and to the trigger delay as obtained from bolus tracking was performed. For statistical analysis, intraclass correlation coefficient and modified variation coefficient (MVC) were used to assess intraindividual agreement of IC and its variation between different time points, respectively. Furthermore, t tests and analysis of variance with Tukey-Kramer post hoc test were used.
RESULTSThe mean intraclass correlation coefficient over all regions of interest was good to excellent (0.642–0.936), irrespective of application of normalization or the normalization technique. Overall, MVC ranged from 1.8% to 25.4%, with significantly lower MVC in data normalized to the aorta (5.8% 1.8%–15.8%) in comparison with the MVC of not normalized data and data normalized to the trigger delay (P < 0.01 and P = 0.04, respectively).
CONCLUSIONSOur study confirms intraindividual, longitudinal variation of DECT-derived IC, which varies among vessels, lymph nodes, organs, and connective tissue, following different perfusion characteristics; normalizing to the aorta seems to improve reproducibility when using a constant contrast media injection protocol.
Purpose Cardiovascular comorbidity anticipates severe progression of COVID-19 and becomes evident by coronary artery calcification (CAC) on low-dose chest computed tomography (LDCT). The purpose of ...this study was to predict a patient's obligation of intensive care treatment by evaluating the coronary calcium burden on the initial diagnostic LDCT. Methods Eighty-nine consecutive patients with parallel LDCT and positive RT-PCR for SARS-CoV-2 were included from three centers. The primary endpoint was admission to ICU, tracheal intubation, or death in the 22-day follow-up period. CAC burden was represented by the Agatston score. Multivariate logistic regression was modeled for prediction of the primary endpoint by the independent variables "Agatston score > 0", as well as the CT lung involvement score, patient sex, age, clinical predictors of severe COVID-19 progression (history of hypertension, diabetes, prior cardiovascular event, active smoking, or hyperlipidemia), and laboratory parameters (creatinine, C-reactive protein, leucocyte, as well as thrombocyte counts, relative lymphocyte count, d-dimer, and lactate dehydrogenase levels). Results After excluding multicollinearity, "Agatston score >0" was an independent regressor within multivariate analysis for prediction of the primary endpoint (p<0.01). Further independent regressors were creatinine (p = 0.02) and leucocyte count (p = 0.04). The Agatston score was significantly higher for COVID-19 cases which completed the primary endpoint (64.2 interquartile range 1.7-409.4 vs. 0 interquartile range 0-0). Conclusion CAC scoring on LDCT might help to predict future obligation of intensive care treatment at the day of patient admission to the hospital.