Quality of referrals for lower extremity ultrasonography and computed tomography pulmonary angiography and associations with positive findings of venous thromboembolism

INTRODUCTION

The referral is the basis for radiologists' assessment of modality, protocol and urgency, and insufficient information may threaten patient safety. The aim of this study was to assess the completeness of referrals for lower extremity venous duplex ultrasonography (LEVDUS) and computed tomography pulmonary angiography (CTPA), and to investigate associations between the provided clinical information including risk factors, symptoms and lab results in the referrals and positive findings of deep vein thrombosis (DVT) and pulmonary embolism (PE), respectively.

METHODS

Referrals for LEVDUS (801) and CTPA (800) performed from 2016 to 2019 were obtained. Three categories of clinical information from the referrals were recorded: symptoms, risk factors and laboratory results, as well as positive imaging findings of venous thromboembolism (VTE). Referral completeness was rated from zero to three according to how many categories of clinical information the referral provided.

RESULTS

Information from all three clinical information categories was provided in 15% and 25% of referrals for LEVDUS and CTPA, respectively, while 2% and 10% of referrals did not contain any clinical information. Symptoms were provided most often (85% for LEVDUS and 94% for CTPA). Provided information about risk factors was significantly associated with positive findings for LEVDUS, (p = 0.02) and CTPA (p < 0.001).

CONCLUSION

A great majority of referrals failed to provide one or more categories of clinical information. Risk factors were associated with a positive finding of VTE on LEVDUS and CTPA.

IMPLICATIONS FOR PRACTICE

Improving clinical information in referrals may improve justification, patient safety and quality of radiology services.

A critical analysis of deficiencies in the quality of information contained in prostate multiparametric MRI requests and reports

BACKGROUND

Multiparametric magnetic resonance imaging (mpMRI) has been increasingly recognised as an important tool in the diagnosis of prostate cancer. PI-RADSv2 guidelines recommend that important clinical information including prostate-specific antigen (PSA) levels, examination findings, and biopsy information should be included in mpMRI requests. PIRADS score and PSA density (PSAD) are both independent predictors for the presence of a clinically significant prostate cancer.

AIMS

This study aims to evaluate the quality of mpMRI requests and reports at our institution in accordance with these parameters.

METHODS

All prostate mpMRIs performed by radiology services in Galway University Hospital between 1st September 2019 and 1st March 2020 were reviewed. Exclusion criteria were applied. Requests and reports were analysed for the presence of the following parameters: PSA-results, examination findings, biopsy information, PI-RADS score, prostate volume, and PSAD.

RESULTS

A total of 586 mpMRIs were performed, and of these, 546 were included. PSA value was provided in 497 (91%) of requests, exam findings in 355 (65%), and biopsy information in 452 (82%). PIRADS score was included in 224 (41%) of reports, prostate volume in 178 (32.6%), and PSAD in 106 (19%).

CONCLUSIONS

Great variation in the quality of information contained in both requests and reports for prostate mpMRIs exists within our service. We aim to improve this by collaborating with our radiology colleagues to develop a proforma for requesting and reporting of mpMRIs for our radiology systems to ensure important clinical and radiological information is provided in future.

Predicting conversion from MCI to AD by integration of rs-fMRI and clinical information using 3D-convolutional neural network

PURPOSE

Alzheimer's is the most common irreversible neurodegenerative disease. Its symptoms range from memory impairments to degradation of multiple cognitive abilities and ultimately death. Mild cognitive impairment (MCI) is the earliest detectable stage that happens between normal aging and early dementia, and even though MCI subjects have a chance of changing back to cognitively normal or even staying the same, there is a risk that their condition progresses to Alzheimer's disease (AD) annually. Therefore predicting AD among MCI subjects is pivotal for starting treatments at an opportune time in case of progression, and if staying stable is the case, the need for consistent medical observations would eliminate. Thus, we aim to diagnose possible conversion from MCI to AD by exploiting a class of deep learning (DL) methods called convolutional neural network (CNN).

METHODS

We proposed a three-dimensional CNN (3D-CNN) to combine and analyze resting-state functional magnetic resonance imaging (rs-fMRI), clinical assessment results, and demographic information to predict conversion from MCI to AD in an average 5-years interval. Initially, a 3D-CNN was developed based on fMRI single volumes of 266 samples from 81 subjects; then, we used neuron layers to combine clinical data with fMRI to improve the results.

RESULTS

At first, the CNN model demonstrated an AUC of 87.67% and an accuracy of 85.7%, then after combining clinical and rs-fMRI features, we observed the following improved scores: an AUC of 91.72%, an accuracy of 87.6%, a sensitivity of 75.58% and a specificity of 92.57%.

CONCLUSION

Our developed algorithm managed to predict prognosis from MCI to AD with high levels of accuracy, proving the potential of DL approaches in solving the matter and the efficiency of integrating clinical information with imaging according to the proposed method.

Moving Radiology Workflow to the Electronic Health Record: Quantitative and Qualitative Experience From a Large Academic Medical Center

RATIONALE AND OBJECTIVES

To objectively and subjectively evaluate a large, academic radiology department's transition to electronic health record (EHR) centered workflow.

MATERIALS AND METHODS

Multiple metrics were compared from before and after the move to EHR-driven workflow. Examination ordering and reading priority data were obtained for 30 days both before and after the transition. Sixteen radiologists were observed opening a computed tomography (CT) examination, and time to open, mouse clicks, and keystrokes were recorded. Information available to the radiologist during interpretation was also compared. Additionally, a 12 question survey was sent out to the residents and faculty both before and after the transition.

RESULTS

Implementation of an eight-level reading priority system increased worklist granularity and improved identification of more urgent studies to read. Radiologists opened CT studies in picture archiving and communications system-driven workflow in 52.4 ± 16.9 seconds using 9.5 ± 3.9 clicks and 6.3 ± 2.9 keystrokes, compared to 17.3 ± 9.5 seconds, 4.8 ± 1.5 clicks, and 0.1 ± 0.3 keystrokes in EHR-driven workflow (p < 0.001 for each measure). More information was available to the radiologist during examination interpretation, and 54.7% of radiologists rated the ease of use of the new system as good or very good (compared to 4.2% for the old system, p < 0.001).

CONCLUSION

Transitioning to an EHR-driven workflow at a large academic medical center improved efficiency, was favorable to radiologists, and enhanced examination prioritization.

The impact of patient clinical information on automated skin cancer detection

Skin cancer is one of the most common types of cancer worldwide. Over the past few years, different approaches have been proposed to deal with automated skin cancer detection. Nonetheless, most of them are based only on dermoscopic images and do not take into account the patient clinical information, an important clue towards clinical diagnosis. In this work, we present an approach to fill this gap. First, we introduce a new dataset composed of clinical images, collected using smartphones, and clinical data related to the patient. Next, we propose a straightforward method that includes an aggregation mechanism in well-known deep learning models to combine features from images and clinical data. Last, we carry out experiments to compare the models' performance with and without using this mechanism. The results present an improvement of approximately 7% in balanced accuracy when the aggregation method is applied. Overall, the impact of clinical data on models' performance is significant and shows the importance of including these features on automated skin cancer detection.

Effect of clinical information and previous exam execution on observer agreement and reliability in the analysis of hysteroscopic video-recordings

INTRODUCTION

Inter-observer agreement and reliability in hysteroscopic image assessment remain uncertain and the type of factors that may influence it has only been studied in relation to the experience of hysteroscopists. We aim to assess the effect of clinical information and previous exam execution on observer agreement and reliability in the analysis of hysteroscopic video-recordings.

MATERIALS AND METHODS

Ninety hysteroscopies were video-recorded and randomized into a group without (Group 1) and with clinical information (Group 2). The videos were independently analyzed by three hysteroscopists, regarding lesion location, dimension, and type, as well as decision to perform a biopsy. One of the hysteroscopists had executed all the exams before. Proportions of agreement (PA) and kappa statistics (κ) with 95% confidence intervals (95% CI) were used.

RESULTS

In Group 2, there was a higher proportion of a normal diagnosis (p < 0.001) and a lower proportion of biopsies recommended (p = 0.027). Observer agreement and reliability were better in Group 2, with the PA and κ ranging, respectively, from 0.73 (95% CI 0.62, 0.83) and 0.44 (95% CI 0.26, 0.63), for image quality, to 0.94 (95% CI 0.88, 0.99) and 0.85 (95% CI 0.65, 0.95), for the decision to perform a biopsy. Execution of the exams before the analysis of the video-recordings did not significantly affect the results.

CONCLUSION

With clinical information, agreement and reliability in the overall analysis of hysteroscopic video-recordings may reach almost perfect results and this was not significantly affected by the execution of the exams before the analysis. However, there is still uncertainty in the analysis of specific endometrial cavity abnormalities.

Impact of patient questionnaires on completeness of clinical information and identification of causes of pain during outpatient abdominopelvic CT interpretation

PURPOSE

To evaluate the impact of questionnaires completed by patients at the time of abdominopelvic CT performed for abdominal pain on the completeness of clinical information and the identification of potential causes of pain, compared with order requisitions alone.

METHODS

100 outpatient CT examinations performed for the evaluation of abdominal pain were retrospectively reviewed. The specificity of the location of pain was compared between the order requisition and patient questionnaire. An abdominal imaging fellow (Reader 1) and abdominal radiologist (Reader 2) reviewed the examinations independently in two sessions 6 weeks apart (one with only the order requisition and one also with the questionnaire). Readers recorded identified causes of pain and rated their confidence in interpretation (1-5 scale; least to greatest confidence).

RESULTS

In 30% of patients, the questionnaire provided a more specific location for pain. Among these, the pain was localized to a specific quadrant in 40%. With having access to the questionnaire, both readers identified additional causes for pain not identified in session 1 (Reader 1, 8.6% [7/81]; Reader 2 5.3% [4/75]). Additional identified causes of pain included diverticulitis, cystitis, peritoneal implants, epiploic appendagitis, osseous metastatic disease, umbilical hernia, gastritis, and SMA syndrome. Confidence in interpretation was significantly greater using the questionnaire for both readers (Reader 1: 4.8 ± 0.6 vs. 4.0 ± 0.5; Reader 2: 4.9 ± 0.3 vs. 4.7 ± 0.5, p < 0.001).

CONCLUSION

Patient questionnaires provide additional relevant clinical history, increased diagnostic yield, and improve radiologists' confidence. Radiology practices are encouraged to implement questionnaires and make these readily available to radiologists at the time of interpretation.

Identifying dynamic pathway interactions based on clinical information

In this paper, we introduce approaches for inferring dynamic pathway interactions by converting static datasets into dynamic datasets using patients' clinical information. One approach uses survival time-based dynamic datasets, and the other uses grade- and stage-based dynamic datasets. Based on cancer grades and stages, we generated six dynamic levels and obtained two pairs of significant pathways out of twelve enriched pathways. One pair of the pathways included CELL ADHESION MOLECULES CAMS and SYSTEMIC LUPUS ERYTHEMATOSUS (correlation coefficient=1.00), in which CD28, CD86, HLA-DOA, and HLA-DOB were identified as common genes in the pathways. The other pair of the pathways included SPLICEOSOME and PRIMARY IMMUNODEFICIENCY (correlation coefficient=0.94) with no common genes identified.

Overview of BioBank Japan follow-up data in 32 diseases

Background

We established a patient-oriented biobank, BioBank Japan, with information on approximately 200,000 patients, suffering from any of 47 common diseases. This follow-up survey focused on 32 diseases, potentially associated with poor vital prognosis, and collected patient survival information, including cause of death. We performed a survival analysis for all subjects to get an overview of BioBank Japan follow-up data.

Methods

A total of 141,612 participants were included. The survival data were last updated in 2014. Kaplan–Meier survival analysis was performed after categorizing subjects according to sex, age group, and disease status. Relative survival rates were estimated using a survival-rate table of the Japanese general population.

Results

Of 141,612 subjects (56.48% male) with 1,087,434 person-years and a 97.0% follow-up rate, 35,482 patients died during follow-up. Mean age at enrollment was 64.24 years for male subjects and 63.98 years for female subjects. The 5-year and 10-year relative survival rates for all subjects were 0.944 and 0.911, respectively, with a median follow-up duration of 8.40 years. Patients with pancreatic cancer had the least favorable prognosis (10-year relative survival: 0.184) and patients with dyslipidemia had the most favorable prognosis (1.013). The most common cause of death was malignant neoplasms. A number of subjects died from diseases other than their registered disease(s).

Conclusions

This is the first report to perform follow-up survival analysis across various common diseases. Further studies should use detailed clinical and genomic information to identify predictors of mortality in patients with common diseases, contributing to the implementation of personalized medicine.

•

141,612 participants with any of 32 diseases were included in the follow-up survey.

•

Subject characteristics at enrollment for the follow-up survey were identified.

•

The relative survival analysis showed the worst prognosis in pancreatic cancer.

•

The most common cause of death in all subjects was malignant neoplasms.

Cross-sectional analysis of BioBank Japan clinical data: A large cohort of 200,000 patients with 47 common diseases

BACKGROUND

To implement personalized medicine, we established a large-scale patient cohort, BioBank Japan, in 2003. BioBank Japan contains DNA, serum, and clinical information derived from approximately 200,000 patients with 47 diseases. Serum and clinical information were collected annually until 2012.

METHODS

We analyzed clinical information of participants at enrollment, including age, sex, body mass index, hypertension, and smoking and drinking status, across 47 diseases, and compared the results with the Japanese database on Patient Survey and National Health and Nutrition Survey. We conducted multivariate logistic regression analysis, adjusting for sex and age, to assess the association between family history and disease development.

RESULTS

Distribution of age at enrollment reflected the typical age of disease onset. Analysis of the clinical information revealed strong associations between smoking and chronic obstructive pulmonary disease, drinking and esophageal cancer, high body mass index and metabolic disease, and hypertension and cardiovascular disease. Logistic regression analysis showed that individuals with a family history of keloid exhibited a higher odds ratio than those without a family history, highlighting the strong impact of host genetic factor(s) on disease onset.

CONCLUSIONS

Cross-sectional analysis of the clinical information of participants at enrollment revealed characteristics of the present cohort. Analysis of family history revealed the impact of host genetic factors on each disease. BioBank Japan, by publicly distributing DNA, serum, and clinical information, could be a fundamental infrastructure for the implementation of personalized medicine.

Predicting the Naturalistic Course of Major Depressive Disorder Using Clinical and Multimodal Neuroimaging Information: A Multivariate Pattern Recognition Study

BACKGROUND

A chronic course of major depressive disorder (MDD) is associated with profound alterations in brain volumes and emotional and cognitive processing. However, no neurobiological markers have been identified that prospectively predict MDD course trajectories. This study evaluated the prognostic value of different neuroimaging modalities, clinical characteristics, and their combination to classify MDD course trajectories.

METHODS

One hundred eighteen MDD patients underwent structural and functional magnetic resonance imaging (MRI) (emotional facial expressions and executive functioning) and were clinically followed-up at 2 years. Three MDD trajectories (chronic n = 23, gradual improving n = 36, and fast remission n = 59) were identified based on Life Chart Interview measuring the presence of symptoms each month. Gaussian process classifiers were employed to evaluate prognostic value of neuroimaging data and clinical characteristics (including baseline severity, duration, and comorbidity).

RESULTS

Chronic patients could be discriminated from patients with more favorable trajectories from neural responses to various emotional faces (up to 73% accuracy) but not from structural MRI and functional MRI related to executive functioning. Chronic patients could also be discriminated from remitted patients based on clinical characteristics (accuracy 69%) but not when age differences between the groups were taken into account. Combining different task contrasts or data sources increased prediction accuracies in some but not all cases.

CONCLUSIONS

Our findings provide evidence that the prediction of naturalistic course of depression over 2 years is improved by considering neuroimaging data especially derived from neural responses to emotional facial expressions. Neural responses to emotional salient faces more accurately predicted outcome than clinical data.

Epithelioid hemangioendothelioma: a study of 14 cytopathology cases

INTRODUCTION

The cytopathologic diagnosis of the rare vascular tumor epithelioid hemangioendothelioma (EHE) in patients who have no previous history of EHE or who have a complicated and/or misleading disease history is challenging. Furthermore, few studies have described the cytopathology of EHE. Herein, we identify 14 cases of EHE from 10 patients, some of whom had a history of epithelial tumor, and provide a detailed report of the cytomorphology of EHE, discuss the tumor's differential diagnoses, and describe ancillary examinations that may be helpful in diagnosing EHE cytologically, especially in patients with a complex disease history.

MATERIALS AND METHODS

We retrieved the slides of 14 cases of EHE archived between 2002 and 2009 in our institution's cytology section. Conventional direct smears and cell block sections were prepared from most fine-needle aspiration samples and from all effusion samples. Cell block sections were subjected to immunostaining for vascular, mesothelial, and epithelial markers.

RESULTS

EHE shared many morphologic features with other, more common tumors such as adenocarcinoma and mesothelioma. The defining cytologic feature of EHE was an intracellular lumen containing entrapped intact and degenerating erythrocytes, which was not present in every case. EHE cells were positive for the vascular markers CD34, CD31, factor VIII, and friend leukemia integration 1 transcription factor (FLI-1) and negative for epithelial and mesothelial markers. Clinicians provided information important to the diagnosis of EHE.

CONCLUSIONS

Carefully examining the smear and cell block sections for morphologic features indicative of EHE (eg, prominent cytoplasmic vacuolization, intranuclear cytoplasmic inclusions, and intracellular lumen containing entrapped intact and degenerating erythrocytes), confirming these findings with immunocytochemical staining, and communicating with clinicians are all important to correctly diagnosing EHE.

Analysis of clinical background in patients with vegetation-like echoes: importance of clinical information

PURPOSE

Detection of vegetation is important for diagnosing infective endocarditis.

METHODS

We analyzed clinical information from 58 patients with vegetation-like echoes on transthoracic echocardiography who had been referred to this institution for an echocardiographic examination during the past 5 years. Patients with healed vegetations were excluded. A vegetation-like echo was defined as a mass, a thread-like echo attached to the valve or endocardium, or both. Diagnosis of a vegetation-like echo required the concurrence of two cardiologists and one sonographer. Altogether, 44 patients were treated with antibiotics because their clinical courses were consistent with active infective endocarditis.

RESULTS

Blood cultures were positive in 27 patients and negative in 17 patients. Follow-up data were available for 10 of the 14 patients who had no findings suggestive of active infective endocarditis. The size of the vegetation-like echo remained unchanged over a mean interval of 12.1 months, and no clinical signs or symptoms of active infective endocarditis appeared. In about one-fourth of the patients with a vegetation-like echo, it was not associated with infective endocarditis.

CONCLUSION

Clinical information, in addition to detection of a vegetation-like echo, appears to be indispensable for diagnosing infective endocarditis.