Optimization of psoriasis assessment system based on patch images.

Psoriasis is a chronic inflammatory skin disease that occurs in various forms throughout the body and is associated with certain conditions such as heart disease, diabetes, and depression. The psoriasis area severity index (PASI) score, a tool used to evaluate the severity of psoriasis, is currently used in clinical trials and clinical research. The determination of severity is based on the subjective judgment of the clinician. Thus, the disease evaluation deviations are induced. Therefore, we propose optimal algorithms that can effectively segment the lesion area and classify the severity. In addition, a new dataset on psoriasis was built, including patch images of erythema and scaling. We performed psoriasis lesion segmentation and classified the disease severity. In addition, we evaluated the best-performing segmentation method and classifier and analyzed features that are highly related to the severity of psoriasis. In conclusion, we presented the optimal techniques for evaluating the severity of psoriasis. Our newly constructed dataset improved the generalization performance of psoriasis diagnosis and evaluation. It proposed an optimal system for specific evaluation indicators of the disease and a quantitative PASI scoring method. The proposed system can help to evaluate the severity of localized psoriasis more accurately.

Automated assessment of the substantia nigra on susceptibility map-weighted imaging using deep convolutional neural networks for diagnosis of Idiopathic Parkinson's disease.

OBJECTIVES: Despite its use in determining nigrostriatal degeneration, the lack of a consistent interpretation of nigrosome 1 susceptibility map-weighted imaging (SMwI) limits its generalized applicability. To implement and evaluate a diagnostic algorithm based on convolutional neural networks for interpreting nigrosome 1 SMwI for determining nigrostriatal degeneration in idiopathic Parkinson's disease (IPD). METHODS: In this retrospective study, we enrolled 267 IPD patients and 160 control subjects (125 patients with drug-induced parkinsonism and 35 healthy subjects) at our institute, and 24 IPD patients and 27 control subjects at three other institutes on approval of the local institutional review boards. Dopamine transporter imaging served as the reference standard for the presence or absence of abnormalities of nigrosome 1 on SMwI. Diagnostic performance was compared between visual assessment by an experienced neuroradiologist and the developed deep learning-based diagnostic algorithm in both internal and external datasets using a bootstrapping method with 10000 re-samples by the "pROC" package of R (version 1.16.2). RESULTS: The area under the receiver operating characteristics curve (AUC) (95% confidence interval [CI]) per participant by the bootstrap method was not significantly different between visual assessment and the deep learning-based algorithm (internal validation, .9622 [0.8912-1.0000] versus 0.9534 [0.8779-0.9956], P = .1511; external validation, 0.9367 [0.8843-0.9802] versus 0.9208 [0.8634-0.9693], P = .6267), indicative of a comparable performance to visual assessment. CONCLUSIONS: Our deep learning-based algorithm for assessing abnormalities of nigrosome 1 on SMwI was found to have a comparable performance to that of an experienced neuroradiologist.

Assessment of a computerized quantitative quality control tool for whole slide images of kidney biopsies.

Inconsistencies in the preparation of histology slides and whole-slide images (WSIs) may lead to challenges with subsequent image analysis and machine learning approaches for interrogating the WSI. These variabilities are especially pronounced in multicenter cohorts, where batch effects (i.e. systematic technical artifacts unrelated to biological variability) may introduce biases to machine learning algorithms. To date, manual quality control (QC) has been the de facto standard for dataset curation, but remains highly subjective and is too laborious in light of the increasing scale of tissue slide digitization efforts. This study aimed to evaluate a computer-aided QC pipeline for facilitating a reproducible QC process of WSI datasets. An open source tool, HistoQC, was employed to identify image artifacts and compute quantitative metrics describing visual attributes of WSIs to the Nephrotic Syndrome Study Network (NEPTUNE) digital pathology repository. A comparison in inter-reader concordance between HistoQC aided and unaided curation was performed to quantify improvements in curation reproducibility. HistoQC metrics were additionally employed to quantify the presence of batch effects within NEPTUNE WSIs. Of the 1814 WSIs (458 H&E, 470 PAS, 438 silver, 448 trichrome) from n = 512 cases considered in this study, approximately 9% (163) were identified as unsuitable for subsequent computational analysis. The concordance in the identification of these WSIs among computational pathologists rose from moderate (Gwet's AC1 range 0.43 to 0.59 across stains) to excellent (Gwet's AC1 range 0.79 to 0.93 across stains) agreement when aided by HistoQC. Furthermore, statistically significant batch effects (p < 0.001) in the NEPTUNE WSI dataset were discovered. Taken together, our findings strongly suggest that quantitative QC is a necessary step in the curation of digital pathology cohorts. © 2020 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Anatomical assessment of trigeminal nerve tractography using diffusion MRI: A comparison of acquisition b-values and single- and multi-fiber tracking strategies.

BACKGROUND: The trigeminal nerve (TGN) is the largest cranial nerve and can be involved in multiple inflammatory, compressive, ischemic or other pathologies. Currently, imaging-based approaches to identify the TGN mostly rely on T2-weighted magnetic resonance imaging (MRI), which provides localization of the cisternal portion of the TGN where the contrast between nerve and cerebrospinal fluid (CSF) is high enough to allow differentiation. The course of the TGN within the brainstem as well as anterior to the cisternal portion, however, is more difficult to display on traditional imaging sequences. An advanced imaging technique, diffusion MRI (dMRI), enables tracking of the trajectory of TGN fibers and has the potential to visualize anatomical regions of the TGN not seen on T2-weighted imaging. This may allow a more comprehensive assessment of the nerve in the context of pathology. To date, most work in TGN tracking has used clinical dMRI acquisitions with a b-value of 1000 s/mm2 and conventional diffusion tensor MRI (DTI) tractography methods. Though higher b-value acquisitions and multi-tensor tractography methods are known to be beneficial for tracking brain white matter fiber tracts, there have been no studies conducted to evaluate the performance of these advanced approaches on nerve tracking of the TGN, in particular on tracking different anatomical regions of the TGN. OBJECTIVE: We compare TGN tracking performance using dMRI data with different b-values, in combination with both single- and multi-tensor tractography methods. Our goal is to assess the advantages and limitations of these different strategies for identifying the anatomical regions of the TGN. METHODS: We proposed seven anatomical rating criteria including true and false positive structures, and we performed an expert rating study of over 1000 TGN visualizations, as follows. We tracked the TGN using high-quality dMRI data from 100 healthy adult subjects from the Human Connectome Project (HCP). TGN tracking performance was compared across dMRI acquisitions with b = 1000 s/mm2, b = 2000 s/mm2 and b = 3000 s/mm2, using single-tensor (1T) and two-tensor (2T) unscented Kalman filter (UKF) tractography. This resulted in a total of six tracking strategies. The TGN was identified using an anatomical region-of-interest (ROI) selection approach. First, in a subset of the dataset we identified ROIs that provided good TGN tracking performance across all tracking strategies. Using these ROIs, the TGN was then tracked in all subjects using the six tracking strategies. An expert rater (GX) visually assessed and scored each TGN based on seven anatomical judgment criteria. These criteria included the presence of multiple expected anatomical segments of the TGN (true positive structures), specifically branch-like structures, cisternal portion, mesencephalic trigeminal tract, and spinal cord tract of the TGN. False positive criteria included the presence of any fibers entering the temporal lobe, the inferior cerebellar peduncle, or the middle cerebellar peduncle. Expert rating scores were analyzed to compare TGN tracking performance across the six tracking strategies. Intra- and inter-rater validation was performed to assess the reliability of the expert TGN rating result. RESULTS: The TGN was selected using two anatomical ROIs (Meckel's Cave and cisternal portion of the TGN). The two-tensor tractography method had significantly better performance on identifying true positive structures, while generating more false positive streamlines in comparison to the single-tensor tractography method. TGN tracking performance was significantly different across the three b-values for almost all structures studied. Tracking performance was reported in terms of the percentage of subjects achieving each anatomical rating criterion. Tracking of the cisternal portion and branching structure of the TGN was generally successful, with the highest performance of over 98% using two-tensor tractography and b = 1000 or b = 2000. However, tracking the smaller mesencephalic and spinal cord tracts of the TGN was quite challenging (highest performance of 37.5% and 57.07%, using two-tensor tractography with b = 1000 and b = 2000, respectively). False positive connections to the temporal lobe (over 38% of subjects for all strategies) and cerebellar peduncles (100% of subjects for all strategies) were prevalent. High joint probability of agreement was obtained in the inter-rater (on average 83%) and intra-rater validation (on average 90%), showing a highly reliable expert rating result. CONCLUSIONS: Overall, the results of the study suggest that researchers and clinicians may benefit from tailoring their acquisition and tracking methodology to the specific anatomical portion of the TGN that is of the greatest interest. For example, tracking of branching structures and TGN-T2 overlap can be best achieved with a two-tensor model and an acquisition using b = 1000 or b = 2000. In general, b = 1000 and b = 2000 acquisitions provided the best-rated tracking results. Further research is needed to improve both sensitivity and specificity of the depiction of the TGN anatomy using dMRI.

Generalised boundary shift integral for longitudinal assessment of spinal cord atrophy.

Spinal cord atrophy measurements obtained from structural magnetic resonance imaging (MRI) are associated with disability in many neurological diseases and serve as in vivo biomarkers of neurodegeneration. Longitudinal spinal cord atrophy rate is commonly determined from the numerical difference between two volumes (based on 3D surface fitting) or two cross-sectional areas (CSA, based on 2D edge detection) obtained at different time-points. Being an indirect measure, atrophy rates are susceptible to variable segmentation errors at the edge of the spinal cord. To overcome those limitations, we developed a new registration-based pipeline that measures atrophy rates directly. We based our approach on the generalised boundary shift integral (GBSI) method, which registers 2 scans and uses a probabilistic XOR mask over the edge of the spinal cord, thereby measuring atrophy more accurately than segmentation-based techniques. Using a large cohort of longitudinal spinal cord images (610 subjects with multiple sclerosis from a multi-centre trial and 52 healthy controls), we demonstrated that GBSI is a sensitive, quantitative and objective measure of longitudinal spinal cord volume change. The GBSI pipeline is repeatable, reproducible, and provides more precise measurements of longitudinal spinal cord atrophy than segmentation-based methods in longitudinal spinal cord atrophy studies.

Video clip assessment of a salivary gland ultrasound scoring system in Sjögren's syndrome using consensual definitions: an OMERACT ultrasound working group reliability exercise.

OBJECTIVE: To develop ultrasound (US) definitions and a US novel scoring system for major salivary gland (SG) lesions in patients with primary Sjögren's syndrome (pSS) and to test their intrareader and inter-reader reliability using US video clips. METHODS: Twenty-five rheumatologists were subjected to a three-round, web-based Delphi process in order to agree on (1) definitions and scanning procedure of salivary gland ultrasonography (SGUS): parotid, submandibular and sublingual glands (PG, SMG and SLG); (2) definitions for the elementary SGUS lesions in patients with Sjögren's syndrome; (3) scoring system for grading changes. The experts rated the statements on a 1-5 Likert scale. In the second step, SGUS video clips of patients with pSS and non-pSS sicca cases were collected containing various spectrums of disease severity followed by an intrareader and inter-reader reliability exercise. Each video clip was evaluated according to the agreed definitions. RESULTS: Consensual definitions were developed after three Delphi rounds. Among the three selected SGs, US assessment of PGs and SMGs was agreed on. Agreement was reached to score only greyscale lesions and to focus on anechoic/hypoechoic foci in a semiquantitative matter or, if not possible on a qualitatively (present/absent) evaluation of fatty or fibrous lesions. Intrareader reliability for detecting and scoring these lesions was excellent (Cohen's kappa 0.81) and inter-reader reliability was good (Light's kappa 0.66). CONCLUSION: New definitions for developing a novel semiquantitative US score in patients with pSS were developed and tested on video clips. Inter-reader and intrareader reliabilities were good and excellent, respectively.

Variability among observers utilizing the CellSolutions BestCyte Cell Sorter imaging system for the assessment of urinary tract cytology specimens.

INTRODUCTION: Image analysis systems are not currently commonly used for evaluating urinary cytology specimens. We evaluated whether the BestCyte Cell Sorter (CellSolutions, Greensboro, NC) imaging system can reliably identify atypical cells in urinary cytology specimens. METHODS: Fifty-three consecutive urine cytology specimens underwent 2 preparations: one slide using SurePath (SP; BD Diagnostics, Sparks, MD)™ for routine clinical evaluation, and a second slide using the CellSolutions F50 system for analysis by the BestCyte Cell Sorter (BCCS) scanning system. Eight observers reviewed atypical cells flagged by BCCS and assigned a BCCS diagnosis to each of the 53 specimens. The observers also blindly reviewed the SP preparation (when available) and assigned an SP diagnosis. The SP diagnoses given by one "expert" observer was considered as a reference diagnosis. RESULTS: There was fair-to-moderate agreement among observers for identifying any atypia and high-grade atypia (Fleiss kappa: 0.417 and 0.338, respectively) using BCCS. Review of SP preparations had slightly better agreement (Fleiss kappa: 0.558 and 0.564, respectively). Intraobserver agreement between the two methods varied greatly between individuals (Cohen's kappa range: 0.260 to 0.647). When a consensus diagnosis could be reached among the observers for cases with surgical follow-up, the consensus diagnosis was concordant in 11 of 12 instances, with one instance being a one-step discrepancy. CONCLUSIONS: Specimen review by BCCS resulted in slightly greater interobserver variability than review of routine SP preparations. This may have been due to variations in observer experience and comfort with the use of a digital imaging system, which is further suggested by the wide range of intraobserver agreement among individuals.

Shake No Bake: A Homemade Ultrasound Gel Recipe for Low-Resource Settings.

Commercially produced ultrasound coupling gel is often a scarce resource in rural regions of low-income countries that use sonography as their main imaging modality and, when available, may be cost prohibitive. Various homemade gels were created and tested to assure image quality was not compromised. Glucomannan-based gel and guar gum-based gel had superior physical properties in initial testing and showed no substantial difference compared with commercially available ultrasound gel on subject and phantom imaging and analysis (P > .05 using a 1-tailed sign test). Neither gel required heating, attracted insects, damaged ultrasound transducers, stained samples of clothing, or had harmful effects to subjects.

Validity of RECIST Version 1.1 for Response Assessment in Metastatic Cancer: A Prospective, Multireader Study.

Purpose To determine the relationship between target lesion selection with use of Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 and classification of therapeutic response in patients with metastatic cancer undergoing systemic cytotoxic and/or targeted therapies. Materials and Methods This prospective multireader study was conducted between July 2015 and July 2017. Three hundred sixteen consecutive participants with metastatic cancer underwent 932 CT examinations to monitor systemic treatment. CT studies were independently read by three radiologists. Readers identified a maximum of five lesions total (and a maximum of two lesions per organ). Dedicated oncology tumor response software was used. The Fleiss κ statistic was used to analyze interreader agreement in the assignment of individual response classes (complete response, partial response, progressive disease, or stable disease) and in the differentiation between progressive and nonprogressive disease. Results Readers selected the same set of target lesions in 128 of the 316 participants (41%) and selected a different set in 188 (59%). When target lesion selection was concordant, agreement was high (assignment of treatment response category: κ = 0.97; 95% confidence interval [CI]: 0.91, 1.0; differentiation between progressive and nonprogressive disease: κ = 0.98; 95% CI: 0.90, 1.0). When target lesion selection was discordant, agreement was significantly reduced (assignment of treatment response category: κ = 0.58; 95% CI: 0.54, 0.62; differentiation between progressive and nonprogressive disease: κ = 0.6; 95% CI: 0.59, 0.70). With concordant target lesion selection, readers agreed regarding diagnosis of progression in 97.7% of participants (95% CI: 95.4%, 100.0%); with discordant target lesion selection, readers agreed in only 55.3% (95% CI: 47.9%, 62.6%) (P < .01). Conclusion In patients with metastatic cancer undergoing systemic treatment, different cancer sites may appear similarly suitable and thus likely to be selected as target lesions but may yield inconsistent or even conflicting results with Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. This indicates that the current, limited set of target lesions in RECIST 1.1 may not reflect overall tumor load or response to therapy. © RSNA, 2018 See also the editorial by Sosna in this issue.

An applicable method for PET/CT image quality assessment and comparison among three PET/CT systems with similar physical performance in cancer patients.

PURPOSE: This research proposes a method with specific procedure guideline for clinical PET/CT image quality assessment according to physicians' behavior of image interpretation and explore the relationship between image quality and image systems with similar physical performance. METHODS: Clinical PET/CT were divided according to body location: brain, chest, abdomen and pelvic cavity. We explored the lesions and suspicious regions where radiologists concerned most through eye-tracker and behavior observation study to generate an assessment checklist. Fifty-five patients who were statistically consistent in age, weight and height were studied. Thirty-seven were scanned with an experimental scanner A and control systems B or C because their clinical pathways required PET/CT examinations at short intervals, the other 18 were scanned with scanners A and C. The grade of every system's PET, CT and PET/CT image performance on the four parts was calculated by subtraction of mean value and variance between experimental and control systems. RESULTS: The scoring checklist was set for PET, CT and PET/CT images in four parts respectively, and a standard procedure guideline was formulated for assessment. Using assessment criteria, the statistical results objectively reflected certain systems' superiority on certain modalities and certain parts of the body. CONCLUSION: Our criteria for clinical PET/CT image quality assessment and comparison were efficient.

The acceptance of the clinical photographic posture assessment tool (CPPAT).

BACKGROUND: There is a lack of evidence-based quantitative clinical methods to adequately assess posture. Our team developed a clinical photographic posture assessment tool (CPPAT) and implemented this tool in clinical practice to standardize posture assessment. The objectives were to determine the level of acceptance of the CPPAT and to document predictors as well as facilitators of and barriers to the acceptance of this tool by clinicians doing posture re-education. METHODS: This is a prospective study focussing on technology acceptance. Thirty-two clinician participants (physical therapists and sport therapists) received a 3-5 h training workshop explaining how to use the CPPAT. Over a three-month trial, they recorded time-on-task for a complete posture evaluation (photo - and photo-processing). Subsequently, participants rated their acceptance of the tool and commented on facilitators and barriers of the clinical method. RESULTS: Twenty-three clinician participants completed the trial. They took 22 (mean) ± 10 min (SD) for photo acquisition and 36 min ± 19 min for photo-processing. Acceptance of the CPPAT was high. Perceived ease of use was an indirect predictor of intention to use, mediated by perceived usefulness. Analysis time was an indirect predictor, mediated by perceived usefulness, and a marginally significant direct predictor. Principal facilitators were objective measurements, visualization, utility, and ease of use. Barriers were time to do a complete analysis of posture, quality of human-computer interaction, non-automation of posture index calculation and photo transfer, and lack of versatility. CONCLUSION: The CPPAT is perceived as useful and easy to use by clinicians and may facilitate the quantitative analysis of posture. Adapting the user-interface and functionality to quantify posture may facilitate a wider adoption of the tool.

Day of Week, Site of Service, and Patient Complexity Disparities in Musculoskeletal MRI Interpretations by Radiologists Versus Nonradiologists.

OBJECTIVE: Although most musculoskeletal MRI examinations are interpreted by radiologists, some nonradiologists provide interpretations as well. We aimed to study day of week (weekday vs weekend), site of service, and patient complexity differences between radiologists and nonradiologists interpreting lower extremity MRI examinations on Medicare beneficiaries. MATERIALS AND METHODS: Using fee-for-service carrier claims for a 5% sample of Medicare beneficiaries nationally from 2012 through 2014, we identified all lower extremity joint MRI examinations. Services were classified by physician specialty, day of week, and site of service. Charlson comorbidity index (CCI) values were calculated for all patients. Chi-square statistical testing was performed. RESULTS: Of all 125,800 billed lower extremity joint MRI examinations, 118,295 (94.0%) were performed on weekdays and 7505 (6.0%) on weekends. Of the weekday examinations, radiologists interpreted 85,991 (83.3%) and nonradiologists 17,260 (16.7%). Of the weekend examinations, radiologists interpreted 6212 (92.8%) and nonradiologists 485 (7.2%). Of examinations performed in inpatient hospital and emergency department settings, radiologists interpreted 6499 (99.2%) and nonradiologists 51 (0.8%). Of the examinations on the most clinically complex patients (CCI ≥ 3), radiologists interpreted 4228 (90.2%) and nonradiologists 461 (9.8%). All interspecialty differences were statistically significant (p < 0.001). CONCLUSION: In the Medicare population, radiologists interpret most lower extremity joint MRI examinations. Compared with nonradiologists, radiologists disproportionately provide services on weekends, in the highest acuity settings, and on the most clinically complex patients. To promote patient access and minimize disparities, future pay-for-performance metrics should consider temporal, acuity, and complexity parameters.

Automated assessment of levodopa-induced dyskinesia: Evaluating the responsiveness of video-based features.

INTRODUCTION: Technological solutions for quantifying Parkinson's disease (PD) symptoms may provide an objective means to track response to treatment, including side effects such as levodopa-induced dyskinesia. Vision-based systems are advantageous as they do not require physical contact with the body and have minimal instrumentation compared to wearables. We have developed a vision-based system to quantify a change in dyskinesia as reported by patients using 2D videos of clinical assessments during acute levodopa infusions. METHODS: Nine participants with PD completed a total of 16 levodopa infusions, where they were asked to report important changes in dyskinesia (i.e. onset and remission). Participants were simultaneously rated using the UDysRS Part III (from video recordings analyzed post-hoc). Body joint positions and movements were tracked using a state-of-the-art deep learning pose estimation algorithm applied to the videos. 416 features (e.g. kinematics, frequency distribution) were extracted to characterize movements. The sensitivity and specificity of each feature to patient-reported changes in dyskinesia severity was computed and compared with physician-rated results. RESULTS: Features achieved similar or superior performance to the UDysRS for detecting the onset and remission of dyskinesia. The best AUC for detecting onset of dyskinesia was 0.822 and for remission of dyskinesia was 0.958, compared to 0.826 and 0.802 for the UDysRS. CONCLUSIONS: Video-based features may provide an objective means of quantifying the severity of levodopa-induced dyskinesia, and have responsiveness as good or better than the clinically-rated UDysRS. The results demonstrate encouraging evidence for future integration of video-based technology into clinical research and eventually clinical practice.

Cassava flour slurry as a low-cost alternative to commercially available gel for obstetrical ultrasound: a blinded non-inferiority trial comparison of image quality.

OBJECTIVE: To evaluate the quality of ultrasound images obtained with cassava flour slurry (CFS) compared with conventional gel in order to determine objectively whether CFS could be a true low-cost alternative. DESIGN: Blinded non-inferiority trial. SETTING: Obstetrical ultrasound unit in an academic medical centre. POPULATION OR SAMPLE: Women with a singleton pregnancy, undergoing anatomy ultrasounds. METHODS: Thirty pregnant women had standard biometry measures obtained with CFS and conventional gel. Images were compared side-by-side in random order by two blinded sonologists and rated for image resolution, detail and total image quality using a 10-cm visual analogue scale. Ratings were compared using paired t-tests. Participant and sonographer experience was measured using five-point Likert scales. MAIN OUTCOME MEASURES: Image resolution, detail, and total image quality. Participant experience of gel regarding irritation, messiness, and ease of removal. RESULTS: We found no significant difference between perceived image quality obtained with CFS (mean = 6.2, SD = 1.2) and commercial gel (mean = 6.4, SD = 1.2) [t (28) = -1.1; P = 0.3]. Images were not rated significantly differently for either reviewer in any measure, any standardized image or any view of a specific anatomic structure. All five sonographers rated CFS as easy to obtain clear images and easy for patient and machine cleanup. Only one participant reported itching with CFS. CONCLUSIONS: CFS produces comparable image quality to commercial ultrasound gel. The dissemination of these results and the simple CFS recipe could significantly increase access to ultrasound for screening, monitoring and diagnostic purposes in resource-limited settings. FUNDING: This study was internally funded by our department. TWEETABLE ABSTRACT: Low-cost homemade cassava flour slurry creates images equal to commercial ultrasound gel, improving access.

Methodologic Guide for Evaluating Clinical Performance and Effect of Artificial Intelligence Technology for Medical Diagnosis and Prediction.

The use of artificial intelligence in medicine is currently an issue of great interest, especially with regard to the diagnostic or predictive analysis of medical images. Adoption of an artificial intelligence tool in clinical practice requires careful confirmation of its clinical utility. Herein, the authors explain key methodology points involved in a clinical evaluation of artificial intelligence technology for use in medicine, especially high-dimensional or overparameterized diagnostic or predictive models in which artificial deep neural networks are used, mainly from the standpoints of clinical epidemiology and biostatistics. First, statistical methods for assessing the discrimination and calibration performances of a diagnostic or predictive model are summarized. Next, the effects of disease manifestation spectrum and disease prevalence on the performance results are explained, followed by a discussion of the difference between evaluating the performance with use of internal and external datasets, the importance of using an adequate external dataset obtained from a well-defined clinical cohort to avoid overestimating the clinical performance as a result of overfitting in high-dimensional or overparameterized classification model and spectrum bias, and the essentials for achieving a more robust clinical evaluation. Finally, the authors review the role of clinical trials and observational outcome studies for ultimate clinical verification of diagnostic or predictive artificial intelligence tools through patient outcomes, beyond performance metrics, and how to design such studies. © RSNA, 2018.

Echocardiographic Assessment of Mitral Stenosis Orifice Area: A Comparison of a Novel Three-Dimensional Method Versus Conventional Techniques.

BACKGROUND: A comprehensive evaluation of mitral stenosis (MS) severity commonly utilizes two-dimensional (2D) echocardiography techniques. However, the complex three-dimensional (3D) structure of the mitral valve (MV) poses challenges to accurate measurements of its orifice area by 2D imaging modalities. We aimed to assess MS severity by comparing measurements of the MV orifice area using conventional echocardiography methods to 3D orifice area (3DOA), a novel echocardiographic technique which minimizes geometric assumptions. METHODS: Routine 2D and 3D intraoperative transesophageal echocardiographic images from 26 adult cardiac surgery patients with at least moderate rheumatic MS were retrospectively reviewed. Measurements of the MV orifice area obtained by pressure half-time (PHT), proximal isovelocity surface area (PISA), continuity equation, and 3D planimetry were compared to those acquired using 3DOA. RESULTS: MV areas derived by PHT, PISA, continuity equation, 3D planimetry, and 3DOA (mean value ± standard deviation) were 1.12 ± 0.27, 1.03 ± 0.27, 1.16 ± 0.35, 0.97 ± 0.25, and 0.76 ± 0.21 cm, respectively. Areas obtained from the 3DOA method were significantly smaller than areas derived from PHT (mean difference 0.35 cm, P < .0001), PISA (mean difference: 0.28 cm, P = .0002), continuity equation (mean difference: 0.43 cm, P = .0015), and 3D planimetry (mean difference: 0.19 cm, P < .0001). MV 3DOAs also identified a significantly greater percentage of patients with severe MS (88%) compared to PHT (31%, P = .006), PISA (42%, P = .01), and continuity equation (39%, P = .017) but not in comparison to 3D planimetry (62%, P = .165). CONCLUSIONS: Novel measures of the stenotic MV 3DOA in patients with rheumatic heart disease are significantly smaller than calculated values obtained by conventional methods and may be consistent with a higher incidence of severe MS compared to 2D techniques. Further investigation is warranted to determine the clinical relevance of 3D echocardiographic techniques used to measure MV area.

Impact of level of expertise versus the statistical tool on vertebral fracture assessment (VFA) readings in cohort studies.

The present study tested if the accuracy of the VFA reading reproducibility is more affected by the statistical tool used or by the reader's level of expertise in 50 VFA from a population-based cohort, the OstéoLaus study. We found that uniform kappa and instruction reading with the ISCD/IOF VFA reading course both increased the accuracy of the reproducibility. INTRODUCTION: Vertebral fractures (VF) due to osteoporosis are under diagnosed. Screening osteoporosis in the general population allows improving management of fragility fracture. It consists to perform a dual X-ray absorptiometry and a spine X-ray to look at a VF. To reduce the dosage of radiation, prevalent or incident VF could be detected by DXA image. The aim of the present study was to test the reproducibility of vertebral fracture assessment (VFA) readings in a population-based cohort and to explore if the accuracy of the reproducibility is more affected by the statistical tool used or by the reader's level of expertise. METHODS: We calculated the reproducibility of VFA reading by uniform and Cohen's kappa, comparing one expert and one non-expert, before and after an instructional on-line International Society of Clinical Densitometry (ISCD) /International Osteoporosis Foundation (IOF) course on VFA reading. We performed the analysis on 50 VFA from a population-based cohort, the OstéoLaus study. RESULTS: Before the VFA reading course, reproducibility with Cohen's kappa was moderate to poor (0 to 0.520), good with the uniform kappa (0.796 to 0.958). After the course, both Cohen's kappa and uniform kappa statistically increased, ranging from 0.524 to 1.000. CONCLUSIONS: For female population-based cohort studies, we recommend using the uniform kappa and instructing a non-expert reader using the ISCD/IOF VFA reading course to correctly read and evaluate the reproducibility of the VFA reading.

A Methodology for Anatomic Ultrasound Image Diagnostic Quality Assessment.

This paper discusses the methods for the assessment of ultrasound image quality based on our experiences with evaluating new methods for anatomic imaging. It presents a methodology to ensure a fair assessment between competing imaging methods using clinically relevant evaluations. The methodology is valuable in the continuing process of method optimization and guided development of new imaging methods. It includes a three phased study plan covering from initial prototype development to clinical assessment. Recommendations to the clinical assessment protocol, software, and statistical analysis are presented. Earlier uses of the methodology has shown that it ensures validity of the assessment, as it separates the influences between developer, investigator, and assessor once a research protocol has been established. This separation reduces confounding influences on the result from the developer to properly reveal the clinical value. This paper exemplifies the methodology using recent studies of synthetic aperture sequential beamforming tissue harmonic imaging.

Imaging study of using radiopharmaceuticals labeled with cyclotron-produced 99mTc.

Cyclotron-produced 99mTc (CPTc) has been recognized as an attractive and practical substitution of reactor/generator based 99mTc. However, the small amount of 92-98Mo in the irradiation of enriched 100Mo could lead to the production of other radioactive technetium isotopes (Tc-impurities) which cannot be chemically separated. Thus, these impurities could contribute to patient dose and affect image quality. The potential radiation dose caused by these Tc-impurities produced using different targets, irradiation conditions, and corresponding to different injection times have been investigated, leading us to create dose-based limits of these parameters for producing clinically acceptable CPTc. However, image quality has been not considered. The aim of the present work is to provide a comprehensive and quantitative analysis of image quality for CPTc. The impact of Tc-impurities in CPTc on image resolution, background noise, and contrast is investigated by performing both Monte-Carlo simulations and phantom experiments. Various targets, irradiation, and acquisition conditions are employed for investigating the image-based limits of CPTc production parameters. Additionally, the relationship between patient dose and image quality of CPTc samples is studied. Only those samples which meet both dose- and image-based limits should be accepted in future clinical studies.

Comparison of setup accuracy of three different image assessment methods for tangential breast radiotherapy.

INTRODUCTION: To compare the differences in setup errors measured with electronic portal image (EPI) and cone-beam computed tomography (CBCT) in patients undergoing tangential breast radiotherapy (RT). Relationship between setup errors, body mass index (BMI) and breast size was assessed. METHODS: Twenty-five patients undergoing postoperative RT to the breast were consented for this study. Weekly CBCT scans were acquired and retrospectively registered to the planning CT in three dimensions, first using bony anatomy for bony registration (CBCT-B) and again using breast tissue outline for soft tissue registration (CBCT-S). Digitally reconstructed radiographs (DRR) generated from CBCT to simulate EPI were compared to the planning DRR using bony anatomy in the V (parallel to the cranio-caudal axis) and U (perpendicular to V) planes. The systematic (Σ) and random (σ) errors were calculated and correlated with BMI and breast size. RESULTS: The systematic and random errors for EPI (ΣV = 3.7 mm, ΣU = 2.8 mm and σV = 2.9 mm, σU = 2.5) and CBCT-B (ΣV = 3.5 mm, ΣU = 3.4 mm and σV = 2.8 mm, σU = 2.8) were of similar magnitude in the V and U planes. Similarly, the differences in setup errors for CBCT-B and CBCT-S in three dimensions were less than 1 mm. Only CBCT-S setup error correlated with BMI and breast size. CONCLUSIONS: CBCT and EPI show insignificant variation in their ability to detect setup error. These findings suggest no significant differences that would make one modality considered superior over the other and EPI should remain the standard of care for most patients. However, there is a correlation with breast size, BMI and setup error as detected by CBCT-S, justifying the use of CBCT-S for larger patients.