Potential of a machine-learning model for dose optimization in CT quality assurance

Automatic chest computed tomography image noise quantification using deep learning

PURPOSE

This study aimed to develop a deep learning (DL) method for noise quantification for clinical chest computed tomography (CT) images without the need for repeated scanning or homogeneous tissue regions.

METHODS

A comprehensive phantom CT dataset (three dose levels, six reconstruction methods, amounting to 9240 slices) was acquired and used to train a convolutional neural network (CNN) to output an estimate of local image noise standard deviations (SD) from a single CT scan input. The CNN model consisting of seven convolutional layers was trained on the phantom image dataset representing a range of scan parameters and was tested with phantom images acquired in a variety of different scan conditions, as well as publicly available chest CT images to produce clinical noise SD maps.

RESULTS

Noise SD maps predicted by the CNN agreed well with the ground truth both visually and numerically in the phantom dataset (errors of < 5 HU for most scan parameter combinations). In addition, the noise SD estimates obtained from clinical chest CT images were similar to running-average based reference estimates in areas without prominent tissue interfaces.

CONCLUSIONS

Predicting local noise magnitudes without the need for repeated scans is feasible using DL. Our implementation trained with phantom data was successfully applied to open-source clinical data with heterogeneous tissue borders and textures. We suggest that automatic DL noise mapping from clinical patient images could be used as a tool for objective CT image quality estimation and protocol optimization.

Complex Relationship Between Artificial Intelligence and CT Radiation Dose

Concerns over need for CT radiation dose optimization and reduction led to improved scanner efficiency and introduction of several reconstruction techniques and image processing-based software. The latest technologies use artificial intelligence (AI) for CT dose optimization and image quality improvement. While CT dose optimization has and can benefit from AI, variations in scanner technologies, reconstruction methods, and scan protocols can lead to substantial variations in radiation doses and image quality across and within different scanners. These variations in turn can influence performance of AI algorithms being deployed for tasks such as detection, segmentation, characterization, and quantification. We review the complex relationship between AI and CT radiation dose.

Utilization of artificial intelligence approach for prediction of DLP values for abdominal CT scans: A high accuracy estimation for risk assessment

Purpose

This study aimed to evaluate Artificial Neural Network (ANN) modeling to estimate the significant dose length product (DLP) value during the abdominal CT examinations for quality assurance in a retrospective, cross-sectional study.

Methods

The structure of the ANN model was designed considering various input parameters, namely patient weight, patient size, body mass index, mean CTDI volume, scanning length, kVp, mAs, exposure time per rotation, and pitch factor. The aforementioned examination details of 551 abdominal CT scans were used as retrospective data. Different types of learning algorithms such as Levenberg-Marquardt, Bayesian and Scaled-Conjugate Gradient were checked in terms of the accuracy of the training data.

Results

The R-value representing the correlation coefficient for the real system and system output is given as 0.925, 0.785, and 0.854 for the Levenberg-Marquardt, Bayesian, and Scaled-Conjugate Gradient algorithms, respectively. The findings showed that the Levenberg-Marquardt algorithm comprehensively detects DLP values for abdominal CT examinations. It can be a helpful approach to simplify CT quality assurance.

Conclusion

It can be concluded that outcomes of this novel artificial intelligence method can be used for high accuracy DLP estimations before the abdominal CT examinations, where the radiation-related risk factors are high or risk evaluation of multiple CT scans is needed for patients in terms of ALARA. Likewise, it can be concluded that artificial learning methods are powerful tools and can be used for different types of radiation-related risk assessments for quality assurance in diagnostic radiology.

Image Quality Control in Lumbar Spine Radiography Using Enhanced U-Net Neural Networks

Purpose

To standardize the radiography imaging procedure, an image quality control framework using the deep learning technique was developed to segment and evaluate lumbar spine x-ray images according to a defined quality control standard.

Materials and Methods

A dataset comprising anteroposterior, lateral, and oblique position lumbar spine x-ray images from 1,389 patients was analyzed in this study. The training set consisted of digital radiography images of 1,070 patients (800, 798, and 623 images of the anteroposterior, lateral, and oblique position, respectively) and the validation set included 319 patients (200, 205, and 156 images of the anteroposterior, lateral, and oblique position, respectively). The quality control standard for lumbar spine x-ray radiography in this study was defined using textbook guidelines of as a reference. An enhanced encoder-decoder fully convolutional network with U-net as the backbone was implemented to segment the anatomical structures in the x-ray images. The segmentations were used to build an automatic assessment method to detect unqualified images. The dice similarity coefficient was used to evaluate segmentation performance.

Results

The dice similarity coefficient of the anteroposterior position images ranged from 0.82 to 0.96 (mean 0.91 ± 0.06); the dice similarity coefficient of the lateral position images ranged from 0.71 to 0.95 (mean 0.87 ± 0.10); the dice similarity coefficient of the oblique position images ranged from 0.66 to 0.93 (mean 0.80 ± 0.14). The accuracy, sensitivity, and specificity of the assessment method on the validation set were 0.971-0.990 (mean 0.98 ± 0.10), 0.714-0.933 (mean 0.86 ± 0.13), and 0.995-1.000 (mean 0.99 ± 0.12) for the three positions, respectively.

Conclusion

This deep learning-based algorithm achieves accurate segmentation of lumbar spine x-ray images. It provides a reliable and efficient method to identify the shape of the lumbar spine while automatically determining the radiographic image quality.

Patient radiation dose during angiography and embolization for abdominal hemorrhage: the influence of CT angiography, fluoroscopy system, patient and procedural variables

Background

Angiography and embolization (AE) is a lifesaving, high radiation dose procedure for treatment of abdominal arterial hemorrhage (AAH). Interventional radiologists have utilized pre-procedure CT angiography (CTA) and newer fluoroscopic systems in an attempt to reduce radiation dose and procedure time.

Purpose

To study the factors contributing to the radiation dose of AE for AAH and to compare to the reference standard.

Materials and methods

This retrospective single-centre observational cohort study identified 154 consecutive AE procedures in 138 patients (median age 65 years; interquartile range 54–77; 103 men) performed with a C-arm fluoroscopic system (Axiom Artis DTA or Axiom Artis Q (Siemens Healthineers)), between January 2010 and December 2017. Parameters analysed included: demographics, fluoroscopy system, bleeding location, body mass index (BMI), preprocedural CT, air kerma-area product (PKA), reference air kerma (K_a,r), fluoroscopy time (FT) and the number of digital subtraction angiography (DSA) runs. Factors affecting dose were assessed using Mann–Whitney U, Kruskal–Wallis one-way ANOVA and linear regression.

Results

Patients treated with the new angiographic system (NS) had a median PKA, median K_a,r, Q3 PKA and Q3 K_a,r that were 74% (p < 0.0005), 66%(p < 0.0005), 55% and 52% lower respectively than those treated with the old system (OS). This dose reduction was consistent for each bleeding location (upper GI, Lower GI and extraluminal). There was no difference in PKA (p = 0.452), K_a,r (p = 0.974) or FT (p = 0.179), between those who did (n = 137) or did not (n = 17) undergo pre-procedure CTA. Other factors significantly influencing radiation dose were: patient BMI and number of DSA runs. A multivariate model containing these variables accounts for 15.2% of the variance in K_a,r (p < 0.005) and 45.9% of the variance of PKA (p < 0.005).

Conclusion

Radiation dose for AE in AAH is significantly reduced by new fluoroscopic technology. Higher patient body mass index is an independent key parameter affecting patient dose. Radiation dose was not influenced by haemorrhage site or performance of pre-procedure CTA.

Upstream Machine Learning in Radiology

Machine learning (ML) and Artificial intelligence (AI) has the potential to dramatically improve radiology practice at multiple stages of the imaging pipeline. Most of the attention has been garnered by applications focused on improving the end of the pipeline: image interpretation. However, this article reviews how AI/ML can be applied to improve upstream components of the imaging pipeline, including exam modality selection, hardware design, exam protocol selection, data acquisition, image reconstruction, and image processing. A breadth of applications and their potential for impact is shown across multiple imaging modalities, including ultrasound, computed tomography, and MRI.

ALIGNING VIDEO-AND STRUCTURED DATA FOR IMAGING OPTIMISATION

Imaging optimisation can benefit from combining structured data with qualitative data in the form of audio and video recordings. Since video is complex to work with, there is a need to find a workable solution that minimises the additional time investment. The purpose of the paper is to outline a general workflow that can begin to address this issue. What is described is a data management process comprising the three steps of collection, mining and contextualisation. This process offers a way to work systematically and at a large scale without succumbing to the context loss of statistical methods. The proposed workflow effectively combines the video and structured data to enable a new level of insights in the optimisation process.

Big Data Framework Using Spark Architecture for Dose Optimization Based on Deep Learning in Medical Imaging

Deep learning and machine learning provide more consistent tools and powerful functions for recognition, classification, reconstruction, noise reduction, quantification and segmentation in biomedical image analysis. Some breakthroughs. Recently, some applications of deep learning and machine learning for low-dose optimization in computed tomography have been developed. Due to reconstruction and processing technology, it has become crucial to develop architectures and/or methods based on deep learning algorithms to minimize radiation during computed tomography scan inspections. This chapter is an extension work done by Alla et al. in 2020 and explain that work very well. This chapter introduces the deep learning for computed tomography scan low-dose optimization, shows examples described in the literature, briefly discusses new methods for computed tomography scan image processing, and provides conclusions. We propose a pipeline for low-dose computed tomography scan image reconstruction based on the literature. Our proposed pipeline relies on deep learning and big data technology using Spark Framework. We will discuss with the pipeline proposed in the literature to finally derive the efficiency and importance of our pipeline. A big data architecture using computed tomography images for low-dose optimization is proposed. The proposed architecture relies on deep learning and allows us to develop effective and appropriate methods to process dose optimization with computed tomography scan images. The real realization of the image denoising pipeline shows us that we can reduce the radiation dose and use the pipeline we recommend to improve the quality of the captured image.

Multicentric study of patients receiving 50 or 100 mSv in a single day through CT imaging-frequency determination and imaging protocols involved

OBJECTIVES

To assess the magnitude and characterization of CT imaging protocols of patients receiving 50 or 100 mSv in a single day.

METHODS

In this multicentric retrospective study covering up to 279 hospitals from January 2015 to December 2019, the effective dose (E) as estimated by dose management system from dose length product of patients was filtered and grouped into per-day dose bands (≤ 20, > 20-50, > 50-70, > 70-100, > 100-200, > 200 mSv). Information on patient's age and imaging protocol was noted. The data were analyzed to determine the frequency of occurrence in each dose band. Top 20 CT imaging protocols that led to patients with a dose of ≥ 50 mSv in a single acquisition were identified and their relative frequency was estimated.

RESULTS

A total of approx. 4.3 million (4,283,738) CT exams were performed in approx. 3.9 million (3,880,524) patient-days indicating 9.41% had more than one CT exam in a single day. There were 31,058 (0.8%) patient-days with ≥ 50 mSv and 1191 (0.03%) with ≥ 100 mSv. Nearly 1/3^rd patient-days reaching ≥ 50 mSv were of patients aged 50 years or younger. The top 20 CT imaging protocols that led to ≥ 50 mSv in a single day belonged to the body region (chest or abdomen and pelvis) and nearly one-third were angiographic studies.

CONCLUSIONS

In the first study of its kind, we report that patients with 50 mSv+ in a single day or a single exam are not rare. The information on imaging protocols leading to such doses and their frequency has been provided to help develop dose management strategies.

KEY POINTS

• Our study of 4,283,738 CT exams performed in 3,880,524 patient-days indicates 0.8% with 50 mSv+ and 0.03% with 100 mSv+ in a single day. • A total of 9.41% underwent more than one CT exam in a single day; nearly 1/3^rd of those with 50 mSv+ were ≤ 50 years of age. • Identified top 20 CT imaging protocols that led to 50 mSv+ doses in a single exam. All belong to chest or abdomen and pelvis and nearly 1/3^rd were angiographic studies.

Evaluation of Organ Dose and Image Quality Metrics of Pediatric CT Chest-Abdomen-Pelvis (CAP) Examination: An Anthropomorphic Phantom Study

The aim of this study is to investigate the impact of CT acquisition parameter setting on organ dose and its influence on image quality metrics in pediatric phantom during CT examination. The study was performed on 64-slice multidetector CT scanner (MDCT) Siemens Definition AS (Siemens Sector Healthcare, Forchheim, Germany) using various CT CAP protocols (P1–P9). Tube potential for P1, P2, and P3 protocols were fixed at 100 kVp while P4, P5, and P6 were fixed at 80 kVp with used of various reference noise values. P7, P8, and P9 were the modification of P1 with changes on slice collimation, pitch factor, and tube current modulation (TCM), respectively. TLD-100 chips were inserted into the phantom slab number 7, 9, 10, 12, 13, and 14 to represent thyroid, lung, liver, stomach, gonads, and skin, respectively. The image quality metrics, signal to noise ratio (SNR) and contrast to noise ratio (CNR) values were obtained from the CT console. As a result, this study indicates a potential reduction in the absorbed dose up to 20% to 50% along with reducing tube voltage, tube current, and increasing the slice collimation. There is no significant difference (p > 0.05) observed between the protocols and image metrics.

Artificial Intelligence and the Medical Physicist: Welcome to the Machine

Artificial intelligence (AI) is a branch of computer science dedicated to giving machines or computers the ability to perform human-like cognitive functions, such as learning, problem-solving, and decision making. Since it is showing superior performance than well-trained human beings in many areas, such as image classification, object detection, speech recognition, and decision-making, AI is expected to change profoundly every area of science, including healthcare and the clinical application of physics to healthcare, referred to as medical physics. As a result, the Italian Association of Medical Physics (AIFM) has created the “AI for Medical Physics” (AI4MP) group with the aims of coordinating the efforts, facilitating the communication, and sharing of the knowledge on AI of the medical physicists (MPs) in Italy. The purpose of this review is to summarize the main applications of AI in medical physics, describe the skills of the MPs in research and clinical applications of AI, and define the major challenges of AI in healthcare.

CT Dosimetry: What Has Been Achieved and What Remains to Be Done

Radiation dose in computed tomography (CT) has become a hot topic due to an upward trend in the number of CT procedures worldwide and the relatively high doses associated with these procedures. The main aim of this review article is to provide an overview of the most frequently used metrics for CT radiation dose characterization, discuss their strengths and limitations, and present patient dose assessment methods. Computed tomography dosimetry is still based on a CT dose index (CTDI) measured using 100-mm-long pencil ionization chambers and standard dosimetry phantoms (CTDI100). This dose index is easily measured but has important limitations. Computed tomography dose index underestimates the dose generated by modern CT scanners with wide beam collimation. Manufacturers should report corrected CTDI values in the consoles of CT systems. The size-specific dose estimate has been proposed to provide an estimate of the average dose at the center of the scan volume along the z-axis of a CT scan. Size-specific dose estimate is based on CTDI and conversion factors and, therefore, its calculation incorporates uncertainties associated with the measurement of CTDI. Moreover, the calculation of size-specific dose estimate is straightforward only when the tube current modulation is not activated and when the patient body diameter does not change considerably along the z-axis of the scan. Effective dose can be used to provide typical patient dose values from CT examinations, compare dose between modalities, and communicate radiogenic risks. In practice, effective dose has been used incorrectly, for example, to characterize a CT procedure as a low-dose examination. Organ or tissue doses, not effective doses, are required for assessing the probability of cancer induction in exposed individuals. Monte Carlo simulation is a powerful technique to estimate organ and tissue dose from CT. However, vendors should make available to the research community the required information to model the imaging process of their CT scanners. Personalized dosimetry based on Monte Carlo simulation and patient models allows accurate organ dose estimation. However, it is not user friendly and fast enough to be applied routinely. Future research efforts should involve the development of advanced artificial intelligence algorithms to overcome drawbacks associated with the current equipment-specific and patient-specific dosimetry.