Radiomics signature for automatic hydronephrosis detection in unenhanced Low-Dose CT.

PURPOSE: To investigate the diagnostic performance of an automatic pipeline for detection of hydronephrosis on kidney's parenchyma on unenhanced low-dose CT of the abdomen. METHODS: This retrospective study included 95 patients with confirmed unilateral hydronephrosis in an unenhanced low-dose CT of the abdomen. Data were split into training (n = 67) and test (n = 28) cohorts. Both kidneys for each case were included in further analyses, whereas the kidney without hydronephrosis was used as control. Using the training cohort, we developed a pipeline consisting of a deep-learning model for automatic segmentation (a Convolutional Neural Network based on nnU-Net architecture) of the kidney's parenchyma and a radiomics classifier to detect hydronephrosis. The models were assessed using standard classification metrics, such as area under the ROC curve (AUC), sensitivity and specificity, as well as semantic segmentation metrics, including Dice coefficient and Jaccard index. RESULTS: Using manual segmentation of the kidney's parenchyma, hydronephrosis can be detected with an AUC of 0.84, a sensitivity of 75% and a specificity of 82%, a PPV of 81% and a NPV of 77%. Automatic kidney segmentation achieved a mean Dice score of 0.87 and 0.91 for the right and left kidney, respectively. Additionally, automatic segmentation achieved an AUC of 0.83, a sensitivity of 86%, specificity of 64%, PPV of 71%, and NPV of 82%. CONCLUSION: Our proposed radiomics signature using automatic kidney's parenchyma segmentation allows for accurate hydronephrosis detection on unenhanced low-dose CT scans of the abdomen independently of widened renal pelvis. This method could be used in clinical routine to highlight hydronephrosis to radiologists as well as clinicians, especially in patients with concurrent parapelvic cysts and might reduce time and costs associated with diagnosing hydronephrosis.

Radiomics-Based Machine Learning Model for Diagnosis of Acute Pancreatitis Using Computed Tomography.

In the early diagnostic workup of acute pancreatitis (AP), the role of contrast-enhanced CT is to establish the diagnosis in uncertain cases, assess severity, and detect potential complications like necrosis, fluid collections, bleeding or portal vein thrombosis. The value of texture analysis/radiomics of medical images has rapidly increased during the past decade, and the main focus has been on oncological imaging and tumor classification. Previous studies assessed the value of radiomics for differentiating between malignancies and inflammatory diseases of the pancreas as well as for prediction of AP severity. The aim of our study was to evaluate an automatic machine learning model for AP detection using radiomics analysis. Patients with abdominal pain and contrast-enhanced CT of the abdomen in an emergency setting were retrospectively included in this single-center study. The pancreas was automatically segmented using TotalSegmentator and radiomics features were extracted using PyRadiomics. We performed unsupervised hierarchical clustering and applied the random-forest based Boruta model to select the most important radiomics features. Important features and lipase levels were included in a logistic regression model with AP as the dependent variable. The model was established in a training cohort using fivefold cross-validation and applied to the test cohort (80/20 split). From a total of 1012 patients, 137 patients with AP and 138 patients without AP were included in the final study cohort. Feature selection confirmed 28 important features (mainly shape and first-order features) for the differentiation between AP and controls. The logistic regression model showed excellent diagnostic accuracy of radiomics features for the detection of AP, with an area under the curve (AUC) of 0.932. Using lipase levels only, an AUC of 0.946 was observed. Using both radiomics features and lipase levels, we showed an excellent AUC of 0.933 for the detection of AP. Automated segmentation of the pancreas and consecutive radiomics analysis almost achieved the high diagnostic accuracy of lipase levels, a well-established predictor of AP, and might be considered an additional diagnostic tool in unclear cases. This study provides scientific evidence that automated image analysis of the pancreas achieves comparable diagnostic accuracy to lipase levels and might therefore be used in the future in the rapidly growing era of AI-based image analysis.

Diagnostic performance of photon-counting detector CT for differentiation between adrenal adenomas and metastases.

OBJECTIVES: Aim of this study was to assess the value of virtual non-contrast (VNC) reconstructions in differentiating between adrenal adenomas and metastases on a photon-counting detector CT (PCD-CT). MATERIAL AND METHODS: Patients with adrenal masses and contrast-enhanced CT scans in portal venous phase were included. Image reconstructions were performed, including conventional VNC (VNCConv) and PureCalcium VNC (VNCPC), as well as virtual monochromatic images (VMI, 40-90 keV) and iodine maps. We analyzed images using semi-automatic segmentation of adrenal lesions and extracted quantitative data. Logistic regression models, non-parametric tests, Bland-Altman plots, and a random forest classifier were used for statistical analyses. RESULTS: The final study cohort consisted of 90 patients (36 female, mean age 67.8 years [range 39-87]) with adrenal lesions (45 adenomas, 45 metastases). Compared to metastases, adrenal adenomas showed significantly lower CT-values in VNCConv and VNCPC (p = 0.007). Mean difference between VNC and true non-contrast (TNC) was 17.67 for VNCConv and 14.85 for VNCPC. Random forest classifier and logistic regression models both identified VNCConv and VNCPC as the best discriminators. When using 26 HU as the threshold in VNCConv reconstructions, adenomas could be discriminated from metastases with a sensitivity of 86.7% and a specificity of 75.6%. CONCLUSION: VNC algorithms overestimate CT values compared to TNC in the assessment of adrenal lesions. However, they allow a reliable discrimination between adrenal adenomas and metastases and could be used in clinical routine in near future with an increased threshold (e.g., 26 HU). Further (multi-center) studies with larger patient cohorts and standardized protocols are required. CLINICAL RELEVANCE STATEMENT: VNC reconstructions overestimate CT values compared to TNC. Using a different threshold (e.g., 26 HU compared to the established 10 HU), VNC has a high diagnostic accuracy for the discrimination between adrenal adenomas and metastases. KEY POINTS: • Virtual non-contrast reconstructions may be promising tools to differentiate adrenal lesions and might save further diagnostic tests. • The conventional and a new calcium-preserving virtual non-contrast algorithm tend to systematically overestimate CT-values compared to true non-contrast images. • Therefore, increasing the established threshold for true non-contrast images (e.g., 10HU) may help to differentiate between adrenal adenomas and metastases on contrast-enhanced CT.

Optimal conspicuity of pancreatic ductal adenocarcinoma in virtual monochromatic imaging reconstructions on a photon-counting detector CT: comparison to conventional MDCT.

PURPOSE: To analyze the conspicuity of pancreatic ductal adenocarcinoma (PDAC) in virtual monoenergetic images (VMI) on a novel photon-counting detector CT (PCD-CT) in comparison to energy-integrating CT (EID-CT). METHODS: Inclusion criteria comprised initial diagnosis of PDAC (reference standard: histopathological analysis) and standardized contrast-enhanced CT imaging either on an EID-CT or a PCD-CT. Patients were excluded due to different histopathological diagnosis or missing tumor delineation on CT. On the PCD-CT, 40-190 keV VMI reconstructions were generated. Image noise, tumor-to-pancreas ratio (TPR) and contrast-to-noise ratio (CNR) were analyzed by ROI-based measurements in arterial and portal venous contrast phase. Two board-certified radiologist evaluated image quality and tumor delineation at both, EID-CT and PCD-CT (40 and 70 keV). RESULTS: Thirty-eight patients (mean age 70.4 years ± 10.3 [range 45-91], 27 males; PCD-CT: n=19, EID-CT: n=19) were retrospectively included. On the PCD-CT, tumor conspicuity (reflected by low TPR and high CNR) was significantly improved at low-energy VMI series (≤ 70 keV compared to > 70 keV), both in arterial and in portal venous contrast phase (P < 0.001), reaching the maximum at 40 keV. Comparison between PCD-CT and EID-CT showed significantly higher CNR on the PCD-CT in portal venous contrast phase at < 70 keV (P < 0.016). On the PCD-CT, tumor conspicuity was improved in portal venous contrast phase compared to arterial contrast phase especially at the lower end of the VMI spectrum (≤ 70 keV). Qualitative analysis revealed that tumor delineation is improved in 40 keV reconstructions compared to 70 keV reconstructions on a PCD-CT. CONCLUSION: PCD-CT VMI reconstructions (≤ 70 keV) showed significantly improved conspicuity of PDAC in quantitative and qualitative analysis in both, arterial and portal venous contrast phase, compared to EID-CT, which may be important for early detection of tumor tissue in clinical routine. Tumor delineation was superior in portal venous contrast phase compared to arterial contrast phase.

Learn-Morph-Infer: A new way of solving the inverse problem for brain tumor modeling.

Current treatment planning of patients diagnosed with a brain tumor, such as glioma, could significantly benefit by accessing the spatial distribution of tumor cell concentration. Existing diagnostic modalities, e.g. magnetic resonance imaging (MRI), contrast sufficiently well areas of high cell density. In gliomas, however, they do not portray areas of low cell concentration, which can often serve as a source for the secondary appearance of the tumor after treatment. To estimate tumor cell densities beyond the visible boundaries of the lesion, numerical simulations of tumor growth could complement imaging information by providing estimates of full spatial distributions of tumor cells. Over recent years a corpus of literature on medical image-based tumor modeling was published. It includes different mathematical formalisms describing the forward tumor growth model. Alongside, various parametric inference schemes were developed to perform an efficient tumor model personalization, i.e. solving the inverse problem. However, the unifying drawback of all existing approaches is the time complexity of the model personalization which prohibits a potential integration of the modeling into clinical settings. In this work, we introduce a deep learning based methodology for inferring the patient-specific spatial distribution of brain tumors from T1Gd and FLAIR MRI medical scans. Coined as Learn-Morph-Infer, the method achieves real-time performance in the order of minutes on widely available hardware and the compute time is stable across tumor models of different complexity, such as reaction-diffusion and reaction-advection-diffusion models. We believe the proposed inverse solution approach not only bridges the way for clinical translation of brain tumor personalization but can also be adopted to other scientific and engineering domains.

MR-CT multi-atlas registration guided by fully automated brain structure segmentation with CNNs.

PURPOSE: Computed tomography (CT) is widely used to identify anomalies in brain tissues because their localization is important for diagnosis and therapy planning. Due to the insufficient soft tissue contrast of CT, the division of the brain into anatomical meaningful regions is challenging and is commonly done with magnetic resonance imaging (MRI). METHODS: We propose a multi-atlas registration approach to propagate anatomical information from a standard MRI brain atlas to CT scans. This translation will enable a detailed automated reporting of brain CT exams. We utilize masks of the lateral ventricles and the brain volume of CT images as adjuvant input to guide the registration process. Besides using manual annotations to test the registration in a first step, we then verify that convolutional neural networks (CNNs) are a reliable solution for automatically segmenting structures to enhance the registration process. RESULTS: The registration method obtains mean Dice values of 0.92 and 0.99 in brain ventricles and parenchyma on 22 healthy test cases when using manually segmented structures as guidance. When guiding with automatically segmented structures, the mean Dice values are 0.87 and 0.98, respectively. CONCLUSION: Our registration approach is a fully automated solution to register MRI atlas images to CT scans and thus obtain detailed anatomical information. The proposed CNN segmentation method can be used to obtain masks of ventricles and brain volume which guide the registration.

Quantitative evaluation of the influence of multiple MRI sequences and of pathological tissues on the registration of longitudinal data acquired during brain tumor treatment.

Registration methods facilitate the comparison of multiparametric magnetic resonance images acquired at different stages of brain tumor treatments. Image-based registration solutions are influenced by the sequences chosen to compute the distance measure, and the lack of image correspondences due to the resection cavities and pathological tissues. Nonetheless, an evaluation of the impact of these input parameters on the registration of longitudinal data is still missing. This work evaluates the influence of multiple sequences, namely T1-weighted (T1), T2-weighted (T2), contrast enhanced T1-weighted (T1-CE), and T2 Fluid Attenuated Inversion Recovery (FLAIR), and the exclusion of the pathological tissues on the non-rigid registration of pre- and post-operative images. We here investigate two types of registration methods, an iterative approach and a convolutional neural network solution based on a 3D U-Net. We employ two test sets to compute the mean target registration error (mTRE) based on corresponding landmarks. In the first set, markers are positioned exclusively in the surroundings of the pathology. The methods employing T1-CE achieves the lowest mTREs, with a improvement up to 0.8 mm for the iterative solution. The results are higher than the baseline when using the FLAIR sequence. When excluding the pathology, lower mTREs are observable for most of the methods. In the second test set, corresponding landmarks are located in the entire brain volumes. Both solutions employing T1-CE obtain the lowest mTREs, with a decrease up to 1.16 mm for the iterative method, whereas the results worsen using the FLAIR. When excluding the pathology, an improvement is observable for the CNN method using T1-CE. Both approaches utilizing the T1-CE sequence obtain the best mTREs, whereas the FLAIR is the least informative to guide the registration process. Besides, the exclusion of pathology from the distance measure computation improves the registration of the brain tissues surrounding the tumor. Thus, this work provides the first numerical evaluation of the influence of these parameters on the registration of longitudinal magnetic resonance images, and it can be helpful for developing future algorithms.

Enhanced registration of ultrasound volumes by segmentation of resection cavity in neurosurgical procedures.

PURPOSE: Neurosurgeons can have a better understanding of surgical procedures by comparing ultrasound images obtained at different phases of the tumor resection. However, establishing a direct mapping between subsequent acquisitions is challenging due to the anatomical changes happening during surgery. We propose here a method to improve the registration of ultrasound volumes, by excluding the resection cavity from the registration process. METHODS: The first step of our approach includes the automatic segmentation of the resection cavities in ultrasound volumes, acquired during and after resection. We used a convolution neural network inspired by the 3D U-Net. Then, subsequent ultrasound volumes are registered by excluding the contribution of resection cavity. RESULTS: Regarding the segmentation of the resection cavity, the proposed method achieved a mean DICE index of 0.84 on 27 volumes. Concerning the registration of the subsequent ultrasound acquisitions, we reduced the mTRE of the volumes acquired before and during resection from 3.49 to 1.22 mm. For the set of volumes acquired before and after removal, the mTRE improved from 3.55 to 1.21 mm. CONCLUSIONS: We proposed an innovative registration algorithm to compensate the brain shift affecting ultrasound volumes obtained at subsequent phases of neurosurgical procedures. To the best of our knowledge, our method is the first to exclude automatically segmented resection cavities in the registration of ultrasound volumes in neurosurgery.

Segmentation-based registration of ultrasound volumes for glioma resection in image-guided neurosurgery.

PURPOSE: In image-guided surgery for glioma removal, neurosurgeons usually plan the resection on images acquired before surgery and use them for guidance during the subsequent intervention. However, after the surgical procedure has begun, the preplanning images become unreliable due to the brain shift phenomenon, caused by modifications of anatomical structures and imprecisions in the neuronavigation system. To obtain an updated view of the resection cavity, a solution is to collect intraoperative data, which can be additionally acquired at different stages of the procedure in order to provide a better understanding of the resection. A spatial mapping between structures identified in subsequent acquisitions would be beneficial. We propose here a fully automated segmentation-based registration method to register ultrasound (US) volumes acquired at multiple stages of neurosurgery. METHODS: We chose to segment sulci and falx cerebri in US volumes, which remain visible during resection. To automatically segment these elements, first we trained a convolutional neural network on manually annotated structures in volumes acquired before the opening of the dura mater and then we applied it to segment corresponding structures in different surgical phases. Finally, the obtained masks are used to register US volumes acquired at multiple resection stages. RESULTS: Our method reduces the mean target registration error (mTRE) between volumes acquired before the opening of the dura mater and during resection from 3.49 mm (± 1.55 mm) to 1.36 mm (± 0.61 mm). Moreover, the mTRE between volumes acquired before opening the dura mater and at the end of the resection is reduced from 3.54 mm (± 1.75 mm) to 2.05 mm (± 1.12 mm). CONCLUSION: The segmented structures demonstrated to be good candidates to register US volumes acquired at different neurosurgical phases. Therefore, our solution can compensate brain shift in neurosurgical procedures involving intraoperative US data.