The role of trust in the use of artificial intelligence for chemical risk assessment.

Risk assessment of chemicals is a time-consuming process and needs to be optimized to ensure all chemicals are timely evaluated and regulated. This transition could be stimulated by valuable applications of in silico Artificial Intelligence (AI)/Machine Learning (ML) models. However, implementation of AI/ML models in risk assessment is lagging behind. Most AI/ML models are considered 'black boxes' that lack mechanistical explainability, causing risk assessors to have insufficient trust in their predictions. Here, we explore 'trust' as an essential factor towards regulatory acceptance of AI/ML models. We provide an overview of the elements of trust, including technical and beyond-technical aspects, and highlight elements that are considered most important to build trust by risk assessors. The results provide recommendations for risk assessors and computational modelers for future development of AI/ML models, including: 1) Keep models simple and interpretable; 2) Offer transparency in the data and data curation; 3) Clearly define and communicate the scope/intended purpose; 4) Define adoption criteria; 5) Make models accessible and user-friendly; 6) Demonstrate the added value in practical settings; and 7) Engage in interdisciplinary settings. These recommendations should ideally be acknowledged in future developments to stimulate trust and acceptance of AI/ML models for regulatory purposes.

Physiologically-Based Kinetic Modeling of Intravenously Administered Gold (Au) Nanoparticles.

Physiologically-based kinetic (PBK) modeling is a valuable tool to understand the kinetics of nanoparticles (NPs) in vivo. However, estimating PBK parameters remains challenging and commonly requires animal studies. To develop predictive models to estimate PBK parameter values based on NP characteristics, a database containing PBK parameter values and corresponding NP characteristics is needed. As a first step toward this objective, this study estimates PBK parameters for gold NPs (AuNPs) and provides a comparison of two different NPs. Two animal experiments are conducted in which varying doses of AuNPs attached with polyethylene glycol (PEG) are administered intravenously to rats. The resulting Au concentrations are used to estimate PBK model parameters. The parameters are compared with PBK parameters previously estimated for poly(alkyl cyanoacrylate) NPs loaded with cabazitaxel and for LipImage 815. This study shows that a small initial database of PBK parameters collected for three NPs is already sufficient to formulate new hypotheses on NP characteristics that may be predictive of PBK parameter values. Further research should focus on developing a larger database and on developing quantitative models to predict PBK parameter values.

PACEMweb: a tool for aggregate consumer exposure assessment.

BACKGROUND: To ascertain the safe use of chemicals that are used in multiple consumer products, the aggregate human exposure, arising from combined use of multiple consumer products needs to be assessed. OBJECTIVE: In this work the Probabilistic Aggregate Consumer Exposure Model (PACEM) is presented and discussed. PACEM is implemented in the publicly available web tool, PACEMweb, for aggregate consumer exposure assessment. METHODS: PACEM uses a person-oriented simulation method that is based on realistic product usage information obtained in surveys from several European countries. PACEM evaluates aggregate exposure in a population considering individual use and co-use patterns as well as variation in product composition. Product usage data is included on personal care products (PCPs) and household cleaning products (HCPs). RESULTS: PACEM has been implemented in a web tool that supports broad use in research as well as regulatory risk assessment. PACEM has been evaluated in a number of applications, testing and illustrating the advantage of the person-oriented modeling method. Also, PACEM assessments have been evaluated by comparing its results with biomonitoring information. SIGNIFICANCE: PACEM enables the assessment of realistic aggregate exposure to chemicals in consumer products. It provides detailed insight into the distribution of exposure in a population as well as products that contribute the most to exposure. This allows for better informed decision making in the risk management of chemicals. IMPACT: Realistic assessment of the total, aggregate exposure of consumers to chemicals in consumer products is necessary to guarantee the safe use of chemicals in these products. PACEMweb provides, for the first time, a publicly available tool to assist in realistic aggregate exposure assessment of consumers to chemicals in consumer products.

Physiologically based pharmacokinetic modeling of intravenously administered nanoformulated substances.

The use of nanobiomaterials (NBMs) is becoming increasingly popular in the field of medicine. To improve the understanding on the biodistribution of NBMs, the present study aimed to implement and parametrize a physiologically based pharmacokinetic (PBPK) model. This model was used to describe the biodistribution of two NBMs after intravenous administration in rats, namely, poly(alkyl cyanoacrylate) (PACA) loaded with cabazitaxel (PACA-Cbz), and LipImage™ 815. A Bayesian parameter estimation approach was applied to parametrize the PBPK model using the biodistribution data. Parametrization was performed for two distinct dose groups of PACA-Cbz. Furthermore, parametrizations were performed three distinct dose groups of LipImage™ 815, resulting in a total of five different parametrizations. The results of this study indicate that the PBPK model can be adequately parametrized using biodistribution data. The PBPK parameters estimated for PACA-Cbz, specifically the vascular permeability, the partition coefficient, and the renal clearance rate, substantially differed from those of LipImage™ 815. This emphasizes the presence of kinetic differences between the different formulations and substances and the need of tailoring the parametrization of PBPK models to the NBMs of interest. The kinetic parameters estimated in this study may help to establish a foundation for a more comprehensive database on NBM-specific kinetic information, which is a first, necessary step towards predictive biodistribution modeling. This effort should be supported by the development of robust in vitro methods to quantify kinetic parameters.

A review on the application of deep learning for CT reconstruction, bone segmentation and surgical planning in oral and maxillofacial surgery.

Computer-assisted surgery (CAS) allows clinicians to personalize treatments and surgical interventions and has therefore become an increasingly popular treatment modality in maxillofacial surgery. The current maxillofacial CAS consists of three main steps: (1) CT image reconstruction, (2) bone segmentation, and (3) surgical planning. However, each of these three steps can introduce errors that can heavily affect the treatment outcome. As a consequence, tedious and time-consuming manual post-processing is often necessary to ensure that each step is performed adequately. One way to overcome this issue is by developing and implementing neural networks (NNs) within the maxillofacial CAS workflow. These learning algorithms can be trained to perform specific tasks without the need for explicitly defined rules. In recent years, an extremely large number of novel NN approaches have been proposed for a wide variety of applications, which makes it a difficult task to keep up with all relevant developments. This study therefore aimed to summarize and review all relevant NN approaches applied for CT image reconstruction, bone segmentation, and surgical planning. After full text screening, 76 publications were identified: 32 focusing on CT image reconstruction, 33 focusing on bone segmentation and 11 focusing on surgical planning. Generally, convolutional NNs were most widely used in the identified studies, although the multilayer perceptron was most commonly applied in surgical planning tasks. Moreover, the drawbacks of current approaches and promising research avenues are discussed.

Efficient high cone-angle artifact reduction in circular cone-beam CT using deep learning with geometry-aware dimension reduction.

High cone-angle artifacts (HCAAs) appear frequently in circular cone-beam computed tomography (CBCT) images and can heavily affect diagnosis and treatment planning. To reduce HCAAs in CBCT scans, we propose a novel deep learning approach that reduces the three-dimensional (3D) nature of HCAAs to two-dimensional (2D) problems in an efficient way. Specifically, we exploit the relationship between HCAAs and the rotational scanning geometry by training a convolutional neural network (CNN) using image slices that were radially sampled from CBCT scans. We evaluated this novel approach using a dataset of input CBCT scans affected by HCAAs and high-quality artifact-free target CBCT scans. Two different CNN architectures were employed, namely U-Net and a mixed-scale dense CNN (MS-D Net). The artifact reduction performance of the proposed approach was compared to that of a Cartesian slice-based artifact reduction deep learning approach in which a CNN was trained to remove the HCAAs from Cartesian slices. In addition, all processed CBCT scans were segmented to investigate the impact of HCAAs reduction on the quality of CBCT image segmentation. We demonstrate that the proposed deep learning approach with geometry-aware dimension reduction greatly reduces HCAAs in CBCT scans and outperforms the Cartesian slice-based deep learning approach. Moreover, the proposed artifact reduction approach markedly improves the accuracy of the subsequent segmentation task compared to the Cartesian slice-based workflow.

Comparison of convolutional neural network training strategies for cone-beam CT image segmentation.

BACKGROUND AND OBJECTIVE: Over the past decade, convolutional neural networks (CNNs) have revolutionized the field of medical image segmentation. Prompted by the developments in computational resources and the availability of large datasets, a wide variety of different two-dimensional (2D) and three-dimensional (3D) CNN training strategies have been proposed. However, a systematic comparison of the impact of these strategies on the image segmentation performance is still lacking. Therefore, this study aimed to compare eight different CNN training strategies, namely 2D (axial, sagittal and coronal slices), 2.5D (3 and 5 adjacent slices), majority voting, randomly oriented 2D cross-sections and 3D patches. METHODS: These eight strategies were used to train a U-Net and an MS-D network for the segmentation of simulated cone-beam computed tomography (CBCT) images comprising randomly-placed non-overlapping cylinders and experimental CBCT images of anthropomorphic phantom heads. The resulting segmentation performances were quantitatively compared by calculating Dice similarity coefficients. In addition, all segmented and gold standard experimental CBCT images were converted into virtual 3D models and compared using orientation-based surface comparisons. RESULTS: The CNN training strategy that generally resulted in the best performances on both simulated and experimental CBCT images was majority voting. When employing 2D training strategies, the segmentation performance can be optimized by training on image slices that are perpendicular to the predominant orientation of the anatomical structure of interest. Such spatial features should be taken into account when choosing or developing novel CNN training strategies for medical image segmentation. CONCLUSIONS: The results of this study will help clinicians and engineers to choose the most-suited CNN training strategy for CBCT image segmentation.

Segmentation of dental cone-beam CT scans affected by metal artifacts using a mixed-scale dense convolutional neural network.

PURPOSE: In order to attain anatomical models, surgical guides and implants for computer-assisted surgery, accurate segmentation of bony structures in cone-beam computed tomography (CBCT) scans is required. However, this image segmentation step is often impeded by metal artifacts. Therefore, this study aimed to develop a mixed-scale dense convolutional neural network (MS-D network) for bone segmentation in CBCT scans affected by metal artifacts. METHOD: Training data were acquired from 20 dental CBCT scans affected by metal artifacts. An experienced medical engineer segmented the bony structures in all CBCT scans using global thresholding and manually removed all remaining noise and metal artifacts. The resulting gold standard segmentations were used to train an MS-D network comprising 100 convolutional layers using far fewer trainable parameters than alternative convolutional neural network (CNN) architectures. The bone segmentation performance of the MS-D network was evaluated using a leave-2-out scheme and compared with a clinical snake evolution algorithm and two state-of-the-art CNN architectures (U-Net and ResNet). All segmented CBCT scans were subsequently converted into standard tessellation language (STL) models and geometrically compared with the gold standard. RESULTS: CBCT scans segmented using the MS-D network, U-Net, ResNet and the snake evolution algorithm demonstrated mean Dice similarity coefficients of 0.87 ± 0.06, 0.87 ± 0.07, 0.86 ± 0.05, and 0.78 ± 0.07, respectively. The STL models acquired using the MS-D network, U-Net, ResNet and the snake evolution algorithm demonstrated mean absolute deviations of 0.44 mm ± 0.13 mm, 0.43 mm ± 0.16 mm, 0.40 mm ± 0.12 mm and 0.57 mm ± 0.22 mm, respectively. In contrast to the MS-D network, the ResNet introduced wave-like artifacts in the STL models, whereas the U-Net incorrectly labeled background voxels as bone around the vertebrae in 4 of the 9 CBCT scans containing vertebrae. CONCLUSION: The MS-D network was able to accurately segment bony structures in CBCT scans affected by metal artifacts.

CT image segmentation of bone for medical additive manufacturing using a convolutional neural network.

BACKGROUND: The most tedious and time-consuming task in medical additive manufacturing (AM) is image segmentation. The aim of the present study was to develop and train a convolutional neural network (CNN) for bone segmentation in computed tomography (CT) scans. METHOD: The CNN was trained with CT scans acquired using six different scanners. Standard tessellation language (STL) models of 20 patients who had previously undergone craniotomy and cranioplasty using additively manufactured skull implants served as "gold standard" models during CNN training. The CNN segmented all patient CT scans using a leave-2-out scheme. All segmented CT scans were converted into STL models and geometrically compared with the gold standard STL models. RESULTS: The CT scans segmented using the CNN demonstrated a large overlap with the gold standard segmentation and resulted in a mean Dice similarity coefficient of 0.92 ±â€¯0.04. The CNN-based STL models demonstrated mean surface deviations ranging between -0.19 mm ±â€¯0.86 mm and 1.22 mm ±â€¯1.75 mm, when compared to the gold standard STL models. No major differences were observed between the mean deviations of the CNN-based STL models acquired using six different CT scanners. CONCLUSIONS: The fully-automated CNN was able to accurately segment the skull. CNNs thus offer the opportunity of removing the current prohibitive barriers of time and effort during CT image segmentation, making patient-specific AM constructs more accesible.