Computed tomography and patient risk: Facts, perceptions and uncertainties

Beyond strong labels: Weakly-supervised learning based on Gaussian pseudo labels for the segmentation of ellipse-like vascular structures in non-contrast CTs

Deep learning-based automated segmentation of vascular structures in preoperative CT angiography (CTA) images contributes to computer-assisted diagnosis and interventions. While CTA is the common standard, non-contrast CT imaging has the advantage of avoiding complications associated with contrast agents. However, the challenges of labor-intensive labeling and high labeling variability due to the ambiguity of vascular boundaries hinder conventional strong-label-based, fully-supervised learning in non-contrast CTs. This paper introduces a novel weakly-supervised framework using the elliptical topology nature of vascular structures in CT slices. It includes an efficient annotation process based on our proposed standards, an approach of generating 2D Gaussian heatmaps serving as pseudo labels, and a training process through a combination of voxel reconstruction loss and distribution loss with the pseudo labels. We assess the effectiveness of the proposed method on one local and two public datasets comprising non-contrast CT scans, particularly focusing on the abdominal aorta. On the local dataset, our weakly-supervised learning approach based on pseudo labels outperforms strong-label-based fully-supervised learning (1.54% of Dice score on average), reducing labeling time by around 82.0%. The efficiency in generating pseudo labels allows the inclusion of label-agnostic external data in the training set, leading to an additional improvement in performance (2.74% of Dice score on average) with a reduction of 66.3% labeling time, where the labeling time remains considerably less than that of strong labels. On the public dataset, the pseudo labels achieve an overall improvement of 1.95% in Dice score for 2D models with a reduction of 68% of the Hausdorff distance for 3D model.

Exploring communication gaps and parental needs during paediatric CT scan risk-benefit dialogue in resource-constrained facilities

BACKGROUND

Clear communication during informed consent is crucial in paediatric computed tomography (CT) procedures, particularly in resource-constrained settings. CT offers valuable diagnostic information but carries potential radiation risks, especially for paediatric patients. Parents play a critical role in decision-making, necessitating thorough risk-benefit discussions. This study aimed to explore parental experiences regarding risk-benefit communication during their children's CT scans in under-resourced healthcare facilities.

METHODS

A qualitative approach with a descriptive design was employed. Semi-structured interviews were conducted with 13 purposefully selected and consenting parents accompanying paediatric patients for CT scans at two public hospitals. Data were analysed using Tesch's eight-step method and ATLAS.ti software.

RESULTS

Participants were parents of children aged 0-10 years (8 males, 5 females), with 11 making their first visit to the CT department. Three main themes emerged: (1) Compromised consenting process, characterised by inadequate explanation of consent and limited risk-benefit communication; (2) Procedural information deficiency, including minimal communication about the procedure and lack of information on examination results; and (3) Preference for improved communication, with parents expressing a desire for comprehensive information and varied opinions on who should disseminate this information. Parents reported feeling uninformed, anxious, and unable to make well-informed decisions due to communication gaps.

CONCLUSIONS

Significant improvements are needed in risk-benefit communication during paediatric CT scans. Healthcare providers should use simplified language, visual aids, and patient-centred discussions to enhance understanding and reduce parental anxiety. Radiographers should allocate sufficient time for discussions, involve referring physicians when necessary, and document the informed consent process thoroughly. Addressing these issues can improve patient experiences and contribute to positive health outcomes in resource-constrained settings.

Artificial intelligence applied to coronary artery calcium scans (AI-CAC) significantly improves cardiovascular events prediction

Coronary artery calcium (CAC) scans contain valuable information beyond the Agatston Score which is currently reported for predicting coronary heart disease (CHD) only. We examined whether new artificial intelligence (AI) applied to CAC scans can predict non-CHD events, including heart failure, atrial fibrillation, and stroke. We applied AI-enabled automated cardiac chambers volumetry and calcified plaque characterization to CAC scans (AI-CAC) of 5830 asymptomatic individuals (52.2% women, age 61.7 ± 10.2 years) in the multi-ethnic study of atherosclerosis during 15 years of follow-up, 1773 CVD events accrued. The AUC at 1-, 5-, 10-, and 15-year follow-up for AI-CAC vs. Agatston score was (0.784 vs. 0.701), (0.771 vs. 0.709), (0.789 vs. 0.712) and (0.816 vs. 0.729) (p < 0.0001 for all), respectively. AI-CAC plaque characteristics, including number, location, density, plus number of vessels, significantly improved CHD prediction in the CAC 1-100 cohort vs. Agatston Score. AI-CAC significantly improved the Agatston score for predicting all CVD events.

Harnessing Artificial Intelligence for the Detection and Management of Colorectal Cancer Treatment

Currently, eight million people in the United States suffer from cancer and it is a major global health concern. Early detection and interventions are urgently needed for all cancers, including colorectal cancer. Colorectal cancer is the third most common type of cancer worldwide. Based on the diagnostic efforts to general awareness and lifestyle choices, it is understandable why colorectal cancer is so prevalent today. There is a notable lack of awareness concerning the impact of this cancer and its connection to lifestyle elements, as well as people sometimes mistaking symptoms for a different gastrointestinal condition. Artificial intelligence (AI) may assist in the early detection of all cancers, including colorectal cancer. The usage of AI has exponentially grown in healthcare through extensive research, and since clinical implementation, it has succeeded in improving patient lifestyles, modernizing diagnostic processes, and innovating current treatment strategies. Numerous challenges arise for patients with colorectal cancer and oncologists alike during treatment. For initial screening phases, conventional methods often result in misdiagnosis. Moreover, after detection, determining the course of which colorectal cancer can sometimes contribute to treatment delays. This article touches on recent advancements in AI and its clinical application while shedding light on why this disease is so common today.

Size-specific dose estimates calculated using patient size measurements from scanned projection radiograph in high-resolution chest computed tomography

INTRODUCTION

Size-specific dose estimates (SSDE) are used to assess patient-specific radiation exposure in Computed Tomography (CT), complementing the volume CT dose index (CTDIvol). This study compared SSDE calculated using patient's lateral size from scan projection radiograph (SPR) with SSDE calculated using water equivalent diameter (Dw) from tomographic images in adult chest high-resolution CT (HRCT).

METHODS

In a single-centre study, the CTDIvol and dose-length product (DLP) were recorded from HRCT dose reports of adult patients. Lateral width (SLat), at the centre of the scan range, from the SPR was measured and the SSDE (SSDER) was calculated using conversion factors related to SLat. Average CT number, area of the slice, and lateral size of the patient (AxLat) were measured on the middle slice. The Dw and SSDE from Dw (SSDEW) were calculated. SSDER and SSDEW were compared using Wilcoxon signed rank test. Correlation between patient size and dosimetry parameters were investigated using Spearman Correlation test with statistical significance at P < 0.05. Bland-Altman plot was also used to test agreement between the two SSDE values.

RESULTS

Median CTDIvol, DLP, SSDER and SSDEW were 11.0 mGy, 372 mGy.cm, 11.6 mGy and 12.9 mGy, respectively. Small but statistically significant differences (P < 0.03) were found between SLat and AxLat as well as between SSDER and SSDEW. Bland-Altman analysis resulted in borderline agreement between SSDE values. Moderate correlations were observed between dosimetry quantities and patient size measurements (ρ > 0.640; P < 0.001). SSDEw showed statistically significant correlation (ρ = 0.587 and P < 0.001) with SSDER.

CONCLUSION

SSDER may be used to assess patients' absorbed radiation dose, before the scan, in adult chest HRCT. The median value of SSDER was about 10% lower than the median value SSDEW. However, the SSDEW should be used after the scan to establish effective dose and radiation risk to the patient.

Advancements and Challenges in the Image-Based Diagnosis of Lung and Colon Cancer: A Comprehensive Review

Image-based diagnosis has become a crucial tool in the identification and management of various cancers, particularly lung and colon cancer. This review delves into the latest advancements and ongoing challenges in the field, with a focus on deep learning, machine learning, and image processing techniques applied to X-rays, CT scans, and histopathological images. Significant progress has been made in imaging technologies like computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET), which, when combined with machine learning and artificial intelligence (AI) methodologies, have greatly enhanced the accuracy of cancer detection and characterization. These advances have enabled early detection, more precise tumor localization, personalized treatment plans, and overall improved patient outcomes. However, despite these improvements, challenges persist. Variability in image interpretation, the lack of standardized diagnostic protocols, unequal access to advanced imaging technologies, and concerns over data privacy and security within AI-based systems remain major obstacles. Furthermore, integrating imaging data with broader clinical information is crucial to achieving a more comprehensive approach to cancer diagnosis and treatment. This review provides valuable insights into the recent developments and challenges in image-based diagnosis for lung and colon cancers, underscoring both the remarkable progress and the hurdles that still need to be overcome to optimize cancer care.

Traumatic brain injuries: a neuropsychological review

The best predictor of functional outcome in victims of traumatic brain injury (TBI) is a neuropsychological evaluation. An exponential growth of research into TBI has focused on diagnosis and treatment. Extant literature lacks a comprehensive neuropsychological review that is simultaneously scholarly and practical. In response, our group included, and went beyond a general overview of TBI's, which commonly include definition, types, severity, and pathophysiology. We incorporate reasons behind the use of particular neuroimaging techniques, as well as the most recent findings on common neuropsychological assessments conducted in TBI cases, and their relationship to outcome. In addition, we include tables outlining estimated recovery trajectories of different age groups, their risk factors and we encompass phenomenological studies, further covering the range of existing-promising tools for cognitive rehabilitation/remediation purposes. Finally, we highlight gaps in current research and directions that would be beneficial to pursue.

Design and Characterisation of a Novel Z-Shaped Inductor-Based Wireless Implantable Sensor for Surveillance of Abdominal Aortic Aneurysm Post-Endovascular Repair

PURPOSE

An abdominal aortic aneurysm (AAA) is a dilation of the aorta over its normal diameter (> 3 cm). The minimally invasive treatment adopted uses a stent graft to be deployed into the aneurysm by a catheter to flow blood through it. However, this approach demands frequent monitoring using imaging modalities that involve radiation and contrast agents. Moreover, the multiple follow-ups are expensive, time-consuming, and resource-demanding for healthcare systems. This study proposes a novel wireless implantable medical sensor (WIMS) to measure the aneurysm growth after the endovascular repair.

METHODS

The proposed sensor is composed of a Z-shaped inductor, similar to a stent ring. The proposed design of the sensor is explored by investigating the inductance, resistance, and quality factor of different possible geometries related to a Z-shaped configuration, such as the height and number of struts. The study is conducted through a combination of numerical simulations and experimental tests, with the assessment being carried out at a frequency of 13.56 MHz.

RESULTS

The results show that a higher number of struts result in higher values of inductance and resistance. On the other hand, the increase in the number of struts decreases the quality factor of the Z-shaped inductor due to the presence of high resistance from the inductor. Moreover, it is observed that the influence of the number of struts present in the Z-shaped inductor tends to decrease for larger radii.

CONCLUSIONS

The numerical and experimental evaluation concludes the ability of the proposed sensor to measure the size of the aneurysm.

Localiser radiographs in CT: Current practice, radiation dose, image quality and clinical applications

INTRODUCTION

Survey or localiser radiographs are integral to CT imaging. However, the diverse functions and roles of the localiser radiograph are often obscure to radiographers and radiologists. This scoping review reports the full scope of localiser radiograph use and function in contemporary CT imaging.

METHODS

A scoping review was performed. A systematic literature search was conducted using four databases: MEDLINE, CINAHL, Emcare and Scopus from January 2013 to December 2023. Data extraction was conducted by two review authors and validated by a third reviewer. Thirty-six studies were included in this review.

RESULTS

Three major themes emerged: radiation dose management, image quality considerations and clinical protocol applications. Specifically, the number, order of selection and directions of localiser radiographs significantly impact patient dose and image quality; which are additionally impacted by off-centre patient positioning, which can influence the accuracy of body size estimates and CT numbers. Finally, the optimal selection of localiser radiographs, including exposure parameters (kVp, mAs), can be a part of clinical task-based imaging protocol optimisation.

CONCLUSIONS

The utilities of localiser radiographs in CT imaging are varied. It is salient that radiographers and radiologists understand their role and the impacts of poor application to ensure that radiation dose is minimised and image quality maximised through correct use. Radiographers and radiologists should also be aware of the impact of poor patient positioning on ACTM function, dose and image quality. Additionally, localiser radiographs should be used for clinical task-based protocol optimisation.

IMPLICATIONS FOR PRACTICE

The number, order of selection, direction, patient off-centring, and exposure parameters must be considered when utilising localiser radiographs as they impact dose, image quality, and protocol applications. It is essential for radiographers and radiologists to understand these impacts.

A comparative analysis of different augmentations for brain images

Deep learning (DL) requires a large amount of training data to improve performance and prevent overfitting. To overcome these difficulties, we need to increase the size of the training dataset. This can be done by augmentation on a small dataset. The augmentation approaches must enhance the model's performance during the learning period. There are several types of transformations that can be applied to medical images. These transformations can be applied to the entire dataset or to a subset of the data, depending on the desired outcome. In this study, we categorize data augmentation methods into four groups: Absent augmentation, where no modifications are made; basic augmentation, which includes brightness and contrast adjustments; intermediate augmentation, encompassing a wider array of transformations like rotation, flipping, and shifting in addition to brightness and contrast adjustments; and advanced augmentation, where all transformation layers are employed. We plan to conduct a comprehensive analysis to determine which group performs best when applied to brain CT images. This evaluation aims to identify the augmentation group that produces the most favorable results in terms of improving model accuracy, minimizing diagnostic errors, and ensuring the robustness of the model in the context of brain CT image analysis.

Exploring Patient Preferences for Information About CT Radiation Exposure: Bridging the Gap Between Patient Preference and Physician Practice

Background

CT scan utilizes ionizing radiation poses a danger to the patient's health. Thus, telling the patient about ionizing radiation would be critical in promoting shared decision-making and improving patient-doctor communication. However, few studies have examined this topic broadly.

Objective

The study was conducted to identify the frequency of physicians informing patients about the radiation risk before ordering a CT scan, as well as to examine the association between patients' demographic characteristics and their awareness of the radiation risks associated with CT scans.

Methods

A cross-sectional study was conducted among 387 patients who had undergone CT scans at a tertiary hospital in Riyadh, Saudi Arabia. Data were collected via phone interviews using a structured questionnaire. Chi-squared tests were employed to assess associations between patients' demographic characteristics and their awareness of CT scan radiation risks.

Results

When examining knowledge, 58% of patients knew that CT involves harmful radiation. This knowledge was significantly associated with higher education level and previous experience with CT scans. Regarding doctors' practice of providing information to patients about the scan, 344 (88.9%) patients indicated that their doctor had explained to them why they needed the scan. Only 28 (7.2%) patients stated that their doctor had mentioned the amount of radiation, and 74 (19.1%) patients indicated that doctors mentioned the risks associated with the radiation of the scan. Almost all patients (96.9%) preferred to be told about why they needed a CT scan.

Conclusion

The vast majority of patients who underwent CT scans did not receive enough information about the harm of the scans. However, most of them preferred to know about this harm.

Seeing the unseen: The role of bioimaging techniques for the diagnostic interventions in intervertebral disc degeneration

Intervertebral Disc Degeneration is a pathophysiological condition that primarily affects the spinal discs, causing back pain and neurological deficits. It is caused by the contribution of several factors such as genetic predisposition, age-related degeneration, and lifestyle choices like obesity and physical activity. Even though there are medications to treat pain, there is a lack of medicines for a complete cure. The main difficulty lies in poor diagnosis of the morphological and functional changes in the disc. With the ever-increasing research on bioimaging techniques, new techniques are being developed and repurposed to evaluate disc shape and composition, and their defects like thinning or deformities on the disc, leading to the proper diagnostic intervention in intervertebral disc degeneration. In this review, we aim to present a comprehensive overview of the imaging techniques used in the pre-clinical and clinical stages for the diagnosis of intervertebral disc degeneration. First, we will discuss about patho-anatomy and the pathophysiology of degenerative disc disease with the significance and a brief description of various dyes and tracers utilized for bioimaging. Then we will shed light on the latest advancements in diagnostic modalities in intervertebral disc degeneration; concluded by an analysis of the repercussions of the methodologies and experimental systems employed in identifying mechanisms and developing therapeutic strategies in intervertebral disc degeneration.

Automatic segmentation of esophageal cancer, metastatic lymph nodes and their adjacent structures in CTA images based on the UperNet Swin network

OBJECTIVE

To create a deep-learning automatic segmentation model for esophageal cancer (EC), metastatic lymph nodes (MLNs) and their adjacent structures using the UperNet Swin network and computed tomography angiography (CTA) images and to improve the effectiveness and precision of EC automatic segmentation and TN stage diagnosis.

METHODS

Attention U-Net, UperNet Swin, UNet++ and UNet were used to train the EC segmentation model to automatically segment the EC, esophagus, pericardium, aorta and MLN from CTA images of 182 patients with postoperative pathologically proven EC. The Dice similarity coefficient (DSC), sensitivity, and positive predictive value (PPV) were used to assess their segmentation effectiveness. The volume of EC was calculated using the segmentation results, and the outcomes and times of automatic and human segmentation were compared. All statistical analyses were completed using SPSS 25.0 software.

RESULTS

Among the four EC autosegmentation models, the UperNet Swin had the best autosegmentation results with a DSC of 0.7820 and the highest values of EC sensitivity and PPV. The esophagus, pericardium, aorta and MLN had DSCs of 0.7298, 0.9664, 0.9496 and 0.5091. The DSCs of the UperNet Swin were 0.6164, 0.7842, 0.8190, and 0.7259 for T1-4 EC. The volume of EC and its adjacent structures between the ground truth and UperNet Swin model were not significantly different.

CONCLUSIONS

The UperNet Swin showed excellent efficiency in autosegmentation and volume measurement of EC, MLN and its adjacent structures in different T stage, which can help to T and N stage diagnose EC and will save clinicians time and energy.

Incidence and Mortality Life-Attributable Risks for Patients Subjected to Recurrent CT Examinations and Cumulative Effective Dose Exceeding 100 mSv

Computed tomography (CT) multi-detector array has been heavily utilized over the past decade. While transforming an individual's diagnosis, the risk of developing pathogenesis as a result remains a concern. The main aim of this institutional cumulative effective dose (CED) review is to highlight the number of adult individuals with a record of CED ≥ 100 mSv over a time span of 5 years. Further, we aim to roughly estimate both incidence and mortality life-attributable risks (LARs) for the shortlisted individuals. CT studies performed over one year, in one dedicated trauma and emergency facility, were retrospectively retrieved and analyzed. Individuals with historical radiological CED ≥ 100 mSv were short-listed. LARs were defined and established based on organ, age and gender. Out of the 4406 CT studies reviewed, 22 individuals were found with CED ≥ 100 mSv. CED varied amongst the short-listed individuals, with the highest CED registered being 223.0 mSv, for a 57-year-old male, cumulated over an average study interval of 46.3 days. The highest median mortality risk was for females, 214 per 100,000 registered for the age group 51-60 years. While certain clinical indications and diseases require close follow-up using radiological examinations, the benefit-to-risk ratio should be carefully considered, particularly when CT is requested.

Artificial Intelligence Chatbots' Understanding of the Risks and Benefits of Computed Tomography and Magnetic Resonance Imaging Scenarios

PURPOSE

Patients may seek online information to better understand medical imaging procedures. The purpose of this study was to assess the accuracy of information provided by 2 popular artificial intelligence (AI) chatbots pertaining to common imaging scenarios' risks, benefits, and alternatives.

METHODS

Fourteen imaging-related scenarios pertaining to computed tomography (CT) or magnetic resonance imaging (MRI) were used. Factors including the use of intravenous contrast, the presence of renal disease, and whether the patient was pregnant were included in the analysis. For each scenario, 3 prompts for outlining the (1) risks, (2) benefits, and (3) alternative imaging choices or potential implications of not using contrast were inputted into ChatGPT and Bard. A grading rubric and a 5-point Likert scale was used by 2 independent reviewers to grade responses. Prompt variability and chatbot context dependency were also assessed.

RESULTS

ChatGPT's performance was superior to Bard's in accurately responding to prompts per Likert grading (4.36 ± 0.63 vs 3.25 ± 1.03 seconds, P < .0001). There was substantial agreement between independent reviewer grading for ChatGPT (κ = 0.621) and Bard (κ = 0.684). Response text length was not statistically different between ChatGPT and Bard (2087 ± 256 characters vs 2162 ± 369 characters, P = .24). Response time was longer for ChatGPT (34 ± 2 vs 8 ± 1 seconds, P < .0001).

CONCLUSIONS

ChatGPT performed superior to Bard at outlining risks, benefits, and alternatives to common imaging scenarios. Generally, context dependency and prompt variability did not change chatbot response content. Due to the lack of detailed scientific reasoning and inability to provide patient-specific information, both AI chatbots have limitations as a patient information resource.

Optimal image quality and radiation doses with optimal tube voltages/currents for pediatric anthropomorphic phantom brains

OBJECTIVE

Using pediatric anthropomorphic phantoms (APs), we aimed to determine the scanning tube voltage/current combinations that could achieve optimal image quality and avoid excessive radiation exposure in pediatric patients.

MATERIALS AND METHODS

A 64-slice scanner was used to scan a standard test phantom to determine the volume CT dose indices (CTDIvol), and three pediatric anthropomorphic phantoms (APs) with highly accurate anatomy and tissue-equivalent materials were studied. These specialized APs represented the average 1-year-old, 5-year-old, and 10-year-old children, respectively. The physical phantoms were constructed with brain tissue-equivalent materials having a density of ρ = 1.07 g/cm3, comprising 22 numbered 2.54-cm-thick sections for the 1-year-old, 26 sections for the 5-year-old, and 32 sections for the 10-year-old. They were scanned to acquire brain CT images and determine the standard deviations (SDs), effective doses (EDs), and contrast-to noise ratios (CNRs). The APs were scanned by 21 combinations of tube voltages/currents (80, 100, or 120 kVp/10, 40, 80, 120, 150, 200, or 250 mA) and rotation time/pitch settings of 1 s/0.984:1.

RESULTS

The optimal tube voltage/current combinations yielding optimal image quality were 80 kVp/80 mA for the 1-year-old AP; 80 kVp/120 mA for the 5-year-old AP; and 80 kVp/150 mA for the 10-year-old AP. Because these scanning tube voltages/currents yielded SDs, respectively, of 12.81, 13.09, and 12.26 HU, along with small EDs of 0.31, 0.34, and 0.31 mSv, these parameters and the induced values were expediently defined as optimal.

CONCLUSIONS

The optimal tube voltages/currents that yielded optimal brain image quality, SDs, CNRs, and EDs herein are novel and essentially important. Clinical translation of these optimal values may allow CT diagnosis with low radiation doses to children's heads.

Development of clinical prediction model to guide the use of CT head scans for non-traumatic Thai patient with seizure: A cross-sectional study

The aim of this study was to develop clinical predictor tools for guiding the use of computed tomography (CT) head scans in non-traumatic Thai patients presented with seizure. A prediction model using a retrospective cross-sectional design was conducted. We recruited adult patients (aged ≥ 18 years) who had been diagnosed with seizures by their physicians and had undergone CT head scans for further investigation. Positive CT head defined as the presence of any new lesion that related to the patient's presented seizure officially reported by radiologist. A total of 9 candidate predictors were preselected. The prediction model was developed using a full multivariable logistic regression with backward stepwise elimination. We evaluated the model's predictive performance in terms of its discriminative ability and calibration via AuROC and calibration plot. The application was then constructed based on final model. A total of 362 patients were included into the analysis which comprising of 71 patients with positive CT head findings and 291 patients with normal results. Six final predictors were identified including: Glasgow coma scale, the presence of focal neurological deficit, history of malignancy, history of CVA, Epilepsy, and the presence of alcohol withdrawal symptom. In terms of discriminative ability, the final model demonstrated excellent performance (AuROC of 0.82 (95% CI: 0.76-0.87)). The calibration plot illustrated a good agreement between observed and predicted risks. This prediction model offers a reliable tool for effectively reduce unnecessary use and instill confidence in supporting physicians in determining the need for CT head scans in non-traumatic patients with seizures.

Demonstrating the use of population level data to investigate trends in the rate, radiation dose and cost of Computed Tomography across clinical groups: Are there any areas of concern?

INTRODUCTION

Increases in computed tomography (CT) use may not always reflect clinical need or improve outcomes. This study aimed to demonstrate how population level data can be used to identify variations in care between patient groups, by analysing system-level changes in CT use around the diagnosis of new conditions.

METHODS

Retrospective repeated cross-sectional observational study using West Australian linked administrative records, including 504,723 adults diagnosed with different conditions in 2006, 2012 and 2015. For 90 days pre/post diagnosis, CT use (any and 2+ scans), effective dose (mSv), lifetime attributable risk (LAR) of cancer incidence and mortality from CT, and costs were assessed.

RESULTS

CT use increased from 209.4 per 1000 new diagnoses in 2006 to 258.0 in 2015; increases were observed for all conditions except neoplasms. Healthcare system costs increased for all conditions but neoplasms and mental disorders. Effective dose increased substantially for respiratory (+2.5 mSv, +23.1%, P < 0.001) and circulatory conditions (+2.1 mSv, +15.4%, P < 0.001). The LAR of cancer incidence and mortality from CT increased for endocrine (incidence +23.4%, mortality +18.0%) and respiratory disorders (+21.7%, +23.3%). Mortality LAR increased for circulatory (+12.1%) and nervous system (+11.0%) disorders. The LAR of cancer incidence and mortality reduced for musculoskeletal system disorders, despite an increase in repeated CT in this group.

CONCLUSIONS

Use and costs increased for most conditions except neoplasms and mental and behavioural disorders. More strategic CT use may have occurred in musculoskeletal conditions, while use and radiation burden increased for respiratory, circulatory and nervous system conditions. Using this high-level approach we flag areas requiring deeper investigation into appropriateness and value of care.

Evaluation of positioning accuracy, radiation dose and image quality: artificial intelligence based automatic versus manual positioning for CT KUB

Background

Recent innovations are making radiology more advanced for patient and patient services. Under the immense burden of radiology practice, Artificial Intelligence (AI) assists in obtaining Computed Tomography (CT) images with less scan time, proper patient placement, low radiation dose (RD), and improved image quality (IQ). Hence, the aim of this study was to evaluate and compare the positioning accuracy, RD, and IQ of AI-based automatic and manual positioning techniques for CT kidney ureters and bladder (CT KUB).

Methods

This prospective study included 143 patients in each group who were referred for computed tomography (CT) KUB examination. Group 1 patients underwent manual positioning (MP), and group 2 patients underwent AI-based automatic positioning (AP) for CT KUB examination. The scanning protocol was kept constant for both the groups. The off-center distance, RD, and quantitative and qualitative IQ of each group were evaluated and compared.

Results

The AP group (9.66±6.361 mm) had significantly less patient off-center distance than the MP group (15.12±9.55 mm). There was a significant reduction in RD in the AP group compared with that in the MP group. The quantitative image noise (IN) was lower, with a higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) in the AP group than in the MP group (p<0.05). Qualitative IQ parameters such as IN, sharpness, and overall IQ also showed significant differences (p< 0.05), with higher scores in the AP group than in the MP group.

Conclusions

The AI-based AP showed higher positioning accuracy with less off-center distance (44%), which resulted in 12% reduction in RD and improved IQ for CT KUB imaging compared with MP.

Ultrasound quantification of pleural effusion volume in supine position: comparison of three model formulae

BACKGROUND

To investigate the accuracy of three model formulae for ultrasound quantification of pleural effusion (PE) volume in patients in supine position.

METHODS

A prospective study including 100 patients with thoracentesis and drainage of PE was conducted. Three model formulae (single section model, two section model and multi-section model) were used to calculate the PE volume. The correlation and consistency analyses between calculated volumes derived from three models and actual PE volume were performed.

RESULTS

PE volumes calculated by three models all showed significant linear correlations with actual PE volume in supine position (all p < 0.001). The reliability of multi-section model in predicting PE volume was significantly higher than that of single section model and slightly higher than that of two section model. When compared with actual drainage volume, the intra-class correlation coefficients (ICCs) of single section model, two section model and multi-section model were 0.72, 0.97 and 0.99, respectively. Significant consistency between calculated PE volumes by using two section model and multi-section model existed for full PE volume range (ICC 0.98).

CONCLUSION

Based on the convenience and accuracy of ultrasound quantification of PE volume, two section model is recommended for pleural effusion assessment in routine clinic, though different model formulae can be selected according to clinical needs.

AI-enabled cardiac chambers volumetry in coronary artery calcium scans (AI-CACTM) predicts heart failure and outperforms NT-proBNP: The multi-ethnic study of Atherosclerosis

INTRODUCTION

Coronary artery calcium (CAC) scans contain useful information beyond the Agatston CAC score that is not currently reported. We recently reported that artificial intelligence (AI)-enabled cardiac chambers volumetry in CAC scans (AI-CAC™) predicted incident atrial fibrillation in the Multi-Ethnic Study of Atherosclerosis (MESA). In this study, we investigated the performance of AI-CAC cardiac chambers for prediction of incident heart failure (HF).

METHODS

We applied AI-CAC to 5750 CAC scans of asymptomatic individuals (52% female, White 40%, Black 26%, Hispanic 22% Chinese 12%) free of known cardiovascular disease at the MESA baseline examination (2000-2002). We used the 15-year outcomes data and compared the time-dependent area under the curve (AUC) of AI-CAC volumetry versus NT-proBNP, Agatston score, and 9 known clinical risk factors (age, gender, diabetes, current smoking, hypertension medication, systolic and diastolic blood pressure, LDL, HDL for predicting incident HF over 15 years.

RESULTS

Over 15 years of follow-up, 256 HF events accrued. The time-dependent AUC [95% CI] at 15 years for predicting HF with AI-CAC all chambers volumetry (0.86 [0.82,0.91]) was significantly higher than NT-proBNP (0.74 [0.69, 0.77]) and Agatston score (0.71 [0.68, 0.78]) (p ​< ​0.0001), and comparable to clinical risk factors (0.85, p ​= ​0.4141). Category-free Net Reclassification Index (NRI) [95% CI] adding AI-CAC LV significantly improved on clinical risk factors (0.32 [0.16,0.41]), NT-proBNP (0.46 [0.33,0.58]), and Agatston score (0.71 [0.57,0.81]) for HF prediction at 15 years (p ​< ​0.0001).

CONCLUSION

AI-CAC volumetry significantly outperformed NT-proBNP and the Agatston CAC score, and significantly improved the AUC and category-free NRI of clinical risk factors for incident HF prediction.