Background: Preoperative planning of patients diagnosed with pancreatic head cancer is difficult and requires specific expertise. This pilot study assesses the added value of three-dimensional (3D) patient models and computer-aided detection (CAD) algorithms in determining the resectability of pancreatic head tumors. Methods: This study included 14 hepatopancreatobiliary experts from eight hospitals. The participants assessed three radiologically resectable and three radiologically borderline resectable cases in a simulated setting via crossover design. Groups were divided in controls (using a CT scan), a 3D group (using a CT scan and 3D models), and a CAD group (using a CT scan, 3D and CAD). For the perceived fulfillment of preoperative needs, the quality and confidence of clinical decision-making were evaluated. Results: A higher perceived ability to determine degrees and the length of tumor–vessel contact was reported in the CAD group compared to controls (p = 0.022 and p = 0.003, respectively). Lower degrees of tumor–vessel contact were predicted for radiologically borderline resectable tumors in the CAD group compared to controls (p = 0.037). Higher confidence levels were observed in predicting the need for vascular resection in the 3D group compared to controls (p = 0.033) for all cases combined. Conclusions: “CAD (including 3D) improved experts’ perceived ability to accurately assess vessel involvement and supports the development of evolving techniques that may enhance the diagnosis and treatment of pancreatic cancer”.

Background Many promising artificial intelligence (AI) and computer-aided detection and diagnosis systems have been developed, but few have been successfully integrated into clinical practice. This is partially owing to a lack of user-centered design of AI-based computer-aided detection or diagnosis (AI-CAD) systems. Objective We aimed to assess the impact of different onboarding tutorials and levels of AI model explainability on radiologists’ trust in AI and the use of AI recommendations in lung nodule assessment on computed tomography (CT) scans. Methods In total, 20 radiologists from 7 Dutch medical centers performed lung nodule assessment on CT scans under different conditions in a simulated use study as part of a 2×2 repeated-measures quasi-experimental design. Two types of AI onboarding tutorials (reflective vs informative) and 2 levels of AI output (black box vs explainable) were designed. The radiologists first received an onboarding tutorial that was either informative or reflective. Subsequently, each radiologist assessed 7 CT scans, first without AI recommendations. AI recommendations were shown to the radiologist, and they could adjust their initial assessment. Half of the participants received the recommendations via black box AI output and half received explainable AI output. Mental model and psychological trust were measured before onboarding, after onboarding, and after assessing the 7 CT scans. We recorded whether radiologists changed their assessment on found nodules, malignancy prediction, and follow-up advice for each CT assessment. In addition, we analyzed whether radiologists’ trust in their assessments had changed based on the AI recommendations. Results Both variations of onboarding tutorials resulted in a significantly improved mental model of the AI-CAD system (informative P=.01 and reflective P=.01). After using AI-CAD, psychological trust significantly decreased for the group with explainable AI output (P=.02). On the basis of the AI recommendations, radiologists changed the number of reported nodules in 27 of 140 assessments, malignancy prediction in 32 of 140 assessments, and follow-up advice in 12 of 140 assessments. The changes were mostly an increased number of reported nodules, a higher estimated probability of malignancy, and earlier follow-up. The radiologists’ confidence in their found nodules changed in 82 of 140 assessments, in their estimated probability of malignancy in 50 of 140 assessments, and in their follow-up advice in 28 of 140 assessments. These changes were predominantly increases in confidence. The number of changed assessments and radiologists’ confidence did not significantly differ between the groups that received different onboarding tutorials and AI outputs. Conclusions Onboarding tutorials help radiologists gain a better understanding of AI-CAD and facilitate the formation of a correct mental model. If AI explanations do not consistently substantiate the probability of malignancy across patient cases, radiologists’ trust in the AI-CAD system can be impaired. Radiologists’ confidence in their assessments was improved by using the AI recommendations.

The early detection of pancreatic ductal adenocarcinoma (PDAC) is essential for optimal treatment of pancreatic cancer patients. We propose a tumor detection framework to improve the detection of pancreatic head tumors on CT scans. In this retrospective research study, CT images of 99 patients with pancreatic head cancer and 98 control cases from the Catharina Hospital Eindhoven were collected. A multi-stage 3D U-Net-based approach was used for PDAC detection including clinically significant secondary features such as pancreatic duct and common bile duct dilation. The developed algorithm was evaluated using a local test set comprising 59 CT scans. The model was externally validated in 28 pancreatic cancer cases of a publicly available medical decathlon dataset. The tumor detection framework achieved a sensitivity of 0.97 and a specificity of 1.00, with an area under the receiver operating curve (AUROC) of 0.99, in detecting pancreatic head cancer in the local test set. In the external test set, we obtained similar results, with a sensitivity of 1.00. The model provided the tumor location with acceptable accuracy obtaining a DICE Similarity Coefficient (DSC) of 0.37. This study shows that a tumor detection framework utilizing CT scans and secondary signs of pancreatic cancer can detect pancreatic tumors with high accuracy.

Artificial Intelligence (AI) decision-making tools for radiology demonstrated potential capacity to improve radiology work in several tasks such as tumor detection. However, relatively low acceptance in clinical practice demonstrates the challenge of incorporating end-users’ lived experience and their opinions to improve the interaction between clinicians and AI solutions. Therefore, we conducted semi-structured interviews with radiologists and technicians who had lived experience with current or the prior generations of radiology AI tools (e.g., Computer Aided Decision tools). Three key themes were elicited. Firstly, the role of AI, addresses how radiology professionals interact with radiology AI; the second theme, adoption in practice, discusses the requirements for easy usage and smooth transition; the third theme, building appropriate trust, explores influencing factors of clinicians’ trust towards radiology AI. Our findings call attention to the adoption of actionable recommendations on the interaction design and the importance of individual tailored functionalities in radiology AI systems.

Lung cancer has both high incidence and mortality rates compared to other cancer types. One important factor for improved patient survival is early detection. Deep learning for lung nodule detection has been extensively studied, as a tool to facilitate clinicians with early nodule detection and classification. Many publications are reporting high detection accuracy and several models have been introduced to clinical practice. However, certain models may have reduced performance in real-world clinical practice. In this study, we introduce a method to assess the robustness of lung nodule detection models. Medically relevant image perturbations are used to assess the robustness of these models. The perturbations include noise and motion perturbations, which have been created in consultation with an expert radiologist to ensure the clinical relevance of the artifacts for thoracic computed tomography (CT) scans. The evaluated models demonstrate robustness to clinically relevant noise simulations, but it shows less resilience to motion artifacts in perturbed CT scans. This robustness evaluation method, incorporating simulated relevant artifacts, can be extended for use in other applications involving the analysis of CT scans.

Radiological imaging plays a crucial role in the detection and treatment of pancreatic ductal adenocarcinoma (PDAC). However, there are several challenges associated with the use of these techniques in daily clinical practice. Determination of the presence or absence of cancer using radiological imaging is difficult and requires specific expertise, especially after neoadjuvant therapy. Early detection and characterization of tumors would potentially increase the number of patients who are eligible for curative treatment. Over the last decades, artificial intelligence (AI)-based computer-aided detection (CAD) has rapidly evolved as a means for improving the radiological detection of cancer and the assessment of the extent of disease. Although the results of AI applications seem promising, widespread adoption in clinical practice has not taken place. This narrative review provides an overview of current radiological CAD systems in pancreatic cancer, highlights challenges that are pertinent to clinical practice, and discusses potential solutions for these challenges.

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive cancer with limited treatment options. This research proposes a workflow and deep learning-based segmentation models to automatically assess tumor-vessel involvement, a key factor in determining tumor resectability. Correct assessment of resectability is vital to determine treatment options. The proposed workflow involves processing CT scans to segment the tumor and vascular structures, analyzing spatial relationships and the extent of vascular involvement, which follows a similar way of working as expert radiologists in PDAC assessment. Three segmentation architectures (nnU-Net, 3D U-Net, and Probabilistic 3D U-Net) achieve a high accuracy in segmenting veins, arteries, and the tumor. The segmentations enable automated detection of tumor involvement with high accuracy (0.88 sensitivity and 0.86 specificity) and automated computation of the degree of tumor-vessel contact. Additionally, due to significant inter-observer variability in these important structures, we present the uncertainty captured by each of the models to further increase insights into the predicted involvement. This result provides clinicians with a clear indication of tumor-vessel involvement and may be used to facilitate more informed decision-making for surgical interventions. The proposed method offers a valuable tool for improving patient outcomes, personalized treatment strategies and survival rates in pancreatic cancer.

To reduce the number of missed or misdiagnosed lung nodules on CT scans by radiologists, many Artificial Intelligence (AI) algorithms have been developed. Some algorithms are currently being implemented in clinical practice, but the question is whether radiologists and patients really benefit from the use of these novel tools. This study aimed to review how AI assistance for lung nodule assessment on CT scans affects the performances of radiologists. We searched for studies that evaluated radiologists' performances in the detection or malignancy prediction of lung nodules with and without AI assistance. Concerning detection, radiologists achieved with AI assistance a higher sensitivity and AUC, while the specificity was slightly lower. Concerning malignancy prediction, radiologists achieved with AI assistance generally a higher sensitivity, specificity and AUC. The radiologists' workflows of using the AI assistance were often only described in limited detail in the papers. As recent studies showed improved performances of radiologists with AI assistance, AI assistance for lung nodule assessment holds great promise. To achieve added value of AI tools for lung nodule assessment in clinical practice, more research is required on the clinical validation of AI tools, impact on follow-up recommendations and ways of using AI tools.

Deep learning-based segmentation algorithms have the potential to expedite the cumbersome clinical task of creating detailed target delineations for disease diagnosis and prognosis. However, these algorithms have yet to be widely adopted in clinical practice, partly because the resulting model segmentations often fall short of the necessary accuracy and robustness that clinical practice demands. This research aims to make AI work in the real world, where domain shift is anticipated and inter-observer variability is inherent to medical practice. While current research aims to design models that can address these challenges, we propose an alternative approach that involves minimal user (clinician) interaction in the segmentation process. By combining the pattern recognition abilities of neural networks with the domain knowledge of clinicians, segmentation predictions can deliver the desired clinical result with little effort on the part of clinicians. To test this approach, we implemented, fine-tuned and compared three state-of-the-art (SOTA) interactive AI (IAI) methods for segmenting six different abdominal organs and pancreatic ductal adenocarcinoma (PDAC), an extremely challenging structure to segment, in CT images. We demonstrate that the fine-tuned RITM (Reviving Iterative Training with Mask Guidance for Interactive Segmentation) method can achieve higher segmentation accuracy than non-interactive SOTA models with as few as three clicks, potentially reducing the time required for treatment planning. Overall, IAI may be an effective method for bridging the gap between what deep learning-based segmentation algorithms have to offer and the high standard that is required for patient care.

BACKGROUND:Decision-making in lung cancer is complex due to a rapidly increasing amount of diagnostic data and treatment options. The need for timely and accurate diagnosis and delivery of care demands high-quality multidisciplinary team (MDT) collaboration and coordination. Clinical decision support systems (CDSSs) can potentially support MDTs in constructing a shared mental model of a patient case. This enables the team to assess the strength and completeness of collected diagnostic data, stratification for the right personalized therapy driven by clinical stage and other treatment-influencing factors, and adapt care management strategies when needed. Current CDSSs often have a suboptimal fit into the decision-making workflow, which hampers their impact in clinical practice.METHODS:A CDSS for multidisciplinary decision-making in lung cancer was designed to support the abovementioned goals through presentation of relevant clinical data in line with existing mental model structures of the MDT members. The CDSS was tested in a simulated multidisciplinary tumor board meeting for primary diagnosis and treatment selection, based on de-identified primary lung cancer cases (n=8). Decision course analysis, eye-tracking data and questionnaires were used to assess the impact of the CDSS on constructing shared mental models to improve the decision-making process and outcome.RESULTS:The CDSS supported the team in their self-correcting capacity for accurate diagnosis and TNM classification. It enabled cross-validation of diagnostic findings, surfaced discordance between diagnostic tests and facilitated cancer staging according the diagnostic evidence, as well as spotting contra-indications for personalized treatment selection.CONCLUSIONS:This study shows the potential of CDSS on clinical decision making, when these systems are properly designed in line with clinical thinking. The presented setup enables assessment of the impact of CDSS design on clinical decision making and optimization of CDSSs to maximize their effect on decision quality and confidence.

Anti‐angiogenic agents combined with chemotherapy is an important strategy for the treatment of solid tumors. However, survival benefit is limited, urging the improvement of combination therapies. We aimed to clarify the effects of vascular endothelial growth factor receptor 2 (VEGFR2) targeting on hemodynamic function and penetration of drugs in esophagogastric adenocarcinoma (EAC). Patient‐derived xenograft (PDX) models of EAC were subjected to long‐term and short‐term treatment with anti‐VEGFR2 therapy followed by chemotherapy injection or multi‐agent dynamic contrast‐enhanced (DCE‐) MRI and vascular casting. Long‐term anti‐VEGFR2‐treated tumors showed a relatively lower flow and vessel density resulting in reduced chemotherapy uptake. On the contrary, short‐term VEGFR2 targeting resulted in relatively higher flow, rapid vasodilation, and improved chemotherapy delivery. Assessment of the extracellular matrix (ECM) revealed that short‐term anti‐angiogenic treatment drastically remodels the tumor stroma by inducing nitric oxide synthesis and hyaluronan degradation, thereby dilating the vasculature and improving intratumoral chemotherapy delivery. These previously unrecognized beneficial effects could not be maintained by long‐term VEGFR2 inhibition. As the identified mechanisms are targetable, they offer direct options to enhance the treatment efficacy of anti‐angiogenic therapy combined with chemotherapy in EAC patients.

Abstract Dynamic contrast–enhanced MRI (DCE-MRI) is a promising technique for assessing the response of tumor vasculature to antivascular therapies. Multiagent DCE-MRI employs a combination of low and high molecular weight contrast agents, which potentially improves the accuracy of estimation of tumor hemodynamic and vascular permeability parameters. In this study, we used multiagent DCE-MRI to assess changes in tumor hemodynamics and vascular permeability after vascular-disrupting therapy. Multiagent DCE-MRI (sequential injection of G5 dendrimer, G2 dendrimer, and Gd-DOTA) was performed in tumor-bearing mice before, 2 and 24 hours after treatment with vascular disrupting agent DMXAA or placebo. Constrained DCE-MRI gamma capillary transit time modeling was used to estimate flow F, blood volume fraction vb, mean capillary transit time tc, bolus arrival time td, extracellular extravascular fraction ve, vascular heterogeneity index α−1 (all identical between agents) and extraction fraction E (reflective of permeability), and transfer constant Ktrans (both agent-specific) in perfused pixels. F, vb, and α−1 decreased at both time points after DMXAA, whereas tc increased. E (G2 and G5) showed an initial increase, after which, both parameters restored. Ktrans (G2 and Gd-DOTA) decreased at both time points after treatment. In the control, placebo-treated animals, only F, tc, and Ktrans Gd-DOTA showed significant changes. Histologic perfused tumor fraction was significantly lower in DMXAA-treated versus control animals. Our results show how multiagent tracer-kinetic modeling can accurately determine the effects of vascular-disrupting therapy by separating simultaneous changes in tumor hemodynamics and vascular permeability. Significance: These findings describe a new approach to measure separately the effects of antivascular therapy on tumor hemodynamics and vascular permeability, which could help more rapidly and accurately assess the efficacy of experimental therapy of this class. Cancer Res; 78(6); 1561–70. ©2018 AACR.

Early evaluation of response to therapy is crucial for selecting the optimal therapeutic follow-up strategy for cancer patients. PDT is a photochemistry-based treatment modality that induces tumor tissue damage by cytotoxic oxygen radicals, generated by a pre-injected photosensitive drug upon light irradiation of tumor tissue. Vascular shutdown is an important mechanism of tumor destruction for most PDT protocols. In this study, we assessed the suitability of Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) to evaluate treatment efficacy within a day after photodynamic therapy (PDT), using the tumor vascular response as a biomarker for treatment success. Methods: DCE-MRI at 7 T was used to measure the micro-vascular status of subcutaneous colon carcinoma tumors before, right after, and 24 h after PDT in mice. Maps of the area under the curve (AUC) of the contrast agent concentration were calculated from the DCE-MRI data. Besides, tracer kinetic parameters including Ktrans were calculated using the standard Tofts-Kermode model. Viability of tumor tissue at 24 h after PDT was assessed by histological analysis. Results: PDT led to drastic decreases in AUC and Ktrans or complete loss of enhancement immediately after treatment, indicating a vascular shutdown in treated tumor regions. Histological analysis demonstrated that the treatment induced extensive necrosis in the tumors. For PDT-treated tumors, the viable tumor fraction showed a strong correlation (ρ ≥ 0.85) with the tumor fraction with Ktrans > 0.05 min-1 right after PDT. The viable tumor fraction also correlated strongly with the enhanced fraction, the average Ktrans, and the fraction with Ktrans > 0.05 min-1 at 24 h after PDT. Images of the viability stained tumor sections were registered to the DCE-MRI data, demonstrating a good spatial agreement between regions with Ktrans > 0.05 min-1 and viable tissue regions. Finally, 3D post-treatment viability detection maps were constructed for the tumors of three mice by applying a threshold (0.05 min-1) to Ktrans at 24 h after PDT. As a proof of principle, these maps were compared to actual tumor progression after one week. Complete tumor response was correctly assessed in one animal, while residual viable tumor tissue was detected in the other two at the locations where residual tumor tissue was observed after one week. Conclusion: This study demonstrates that DCE-MRI is an effective tool for early evaluation of PDT tumor treatment.

Noninvasive preclinical methods for the characterization of myocardial vascular function are crucial to an understanding of the dynamics of ischemic cardiac disease. Ischemic heart disease is associated with myocardial endothelial dysfunction, resulting in leakage of plasma albumin into the extravascular space. These features can be harnessed in a novel noninvasive three-dimensional magnetic resonance imaging method to measure fractional blood volume (fBV) and vascular permeability (permeability-surface area product, PS) using labeled albumin as a blood pool contrast agent. C57BL/6 mice were imaged before and 3days after myocardial infarction (MI). Following the quantification of endogenous myocardial R-1, the dynamics of intravenously injected albumin-based contrast agent, extravasating from permeable myocardial blood vessels, were tracked on short-axis magnetic resonance images of the entire heart. This study successfully discriminated between infarcted and remote regions at 3days post-infarct, based on a reduced fBV and increased PS in the infarcted region. These findings were confirmed using ex vivo fluorescence imaging and histology. We have demonstrated a novel method to quantify blood volume and permeability in the infarcted myocardium, providing an imaging biomarker for the assessment of endothelial dysfunction. This method has the potential to three-dimensionally visualize subtle changes in myocardial permeability and to track endothelial function for longitudinal cardiac studies determining pathophysiological processes during infarct healing.

The development of efficient contrast agents for cell labelling coupled with powerful medical imaging techniques is of great interest for monitoring cell trafficking with potential clinical applications such as organ repair and regenerative medicine. In this paper, functionalised multi-walled carbon nanotubes (MWNTs) were engineered for cell labelling in T1-weighted MRI applications. These sophisticated constructs were covalently functionalised with the gadolinium (Gd) chelating agent, diethylene triamine pentaacetic acid (DTPA), enabling tight attachment of Gd atoms onto the nanotube surface. The resulting Gd-labelled MWNTs were found to be stable over 2weeks in water and mouse serum and high payload of Gd atoms could be loaded onto the nanotubes. The r1 relaxivity of the Gd-MWNTs was a 3-fold higher than of the clinically approved T1 contrast agent Magnevist at a magnetic field strength of 7T. The contrast efficiency, expressed as the r1 relaxivity, of the Gd-MWNTs in Human Umbilical Vein Endothelial cells (HUVEC) was investigated at 7T and was found to be around 6.6mM−1s−1. There was no reduction of the contrast efficiency after internalisation in HUVECs, which was imparted to the ability of carbon nanotubes to translocate the cell membrane.

Objective The aim of this study was to characterize response to photodynamic therapy (PDT) in a mouse cancer model using a multi-parametric quantitative MRI protocol and to identify MR parameters as potential biomarkers for early assessment of treatment outcome. Methods CT26.WT colon carcinoma tumors were grown subcutaneously in the hind limb of BALB/c mice. Therapy consisted of intravenous injection of the photosensitizer Bremachlorin, followed by 10 min laser illumination (200 mW/cm2) of the tumor 6 h post injection. MRI at 7 T was performed at baseline, directly after PDT, as well as at 24 h, and 72 h. Tumor relaxation time constants (T1 and T2) and apparent diffusion coefficient (ADC) were quantified at each time point. Additionally, Gd-DOTA dynamic contrast-enhanced (DCE) MRI was performed to estimate transfer constants (Ktrans) and volume fractions of the extravascular extracellular space (ve) using standard Tofts-Kermode tracer kinetic modeling. At the end of the experiment, tumor viability was characterized by histology using NADH-diaphorase staining. Results The therapy induced extensive cell death in the tumor and resulted in significant reduction in tumor growth, as compared to untreated controls. Tumor T1 and T2 relaxation times remained unchanged up to 24 h, but decreased at 72 h after treatment. Tumor ADC values significantly increased at 24 h and 72 h. DCE-MRI derived tracer kinetic parameters displayed an early response to the treatment. Directly after PDT complete vascular shutdown was observed in large parts of the tumors and reduced uptake (decreased Ktrans) in remaining tumor tissue. At 24 h, contrast uptake in most tumors was essentially absent. Out of 5 animals that were monitored for 2 weeks after treatment, 3 had tumor recurrence, in locations that showed strong contrast uptake at 72 h. Conclusion DCE-MRI is an effective tool for visualization of vascular effects directly after PDT. Endogenous contrast parameters T1, T2, and ADC, measured at 24 to 72 h after PDT, are also potential biomarkers for evaluation of therapy outcome.