Zeitschrift fur Medizinische Physik最新文献

Introduction

Multiple sclerosis (MS) is a complex neurodegenerative disorder that affects the brain and spinal cord. In this study, we applied a deep learning-based approach using the StyleGAN model to explore patterns related to MS and predict disease progression in magnetic resonance images (MRI).

Methods

We trained the StyleGAN model unsupervised using T₁-weighted GRE MR images and diffusion-based ADC maps of MS patients and healthy controls. We then used the trained model to resample MR images from real input data and modified them by manipulations in the latent space to simulate MS progression. We analyzed the resulting simulation-related patterns mimicking disease progression by comparing the intensity profiles of the original and manipulated images and determined the brain parenchymal fraction (BPF).

Results

Our results show that MS progression can be simulated by manipulating MR images in the latent space, as evidenced by brain volume loss on both T₁-weighted and ADC maps and increasing lesion extent on ADC maps.

Conclusion

Overall, this study demonstrates the potential of the StyleGAN model in medical imaging to study image markers and to shed more light on the relationship between brain atrophy and MS progression through corresponding manipulations in the latent space.

Purpose

During the radiation treatment planning process, one of the time-consuming procedures is the final high-resolution dose calculation, which obstacles the wide application of the emerging online adaptive radiotherapy techniques (OLART). There is an urgent desire for highly accurate and efficient dose calculation methods. This study aims to develop a dose super resolution-based deep learning model for fast and accurate dose prediction in clinical practice.

Method

A Multi-stage Dose Super-Resolution Network (MDSR Net) architecture with sparse masks module and multi-stage progressive dose distribution restoration method were developed to predict high-resolution dose distribution using low-resolution data. A total of 340 VMAT plans from different disease sites were used, among which 240 randomly selected nasopharyngeal, lung, and cervix cases were used for model training, and the remaining 60 cases from the same sites for model benchmark testing, and additional 40 cases from the unseen site (breast and rectum) was used for model generalizability evaluation. The clinical calculated dose with a grid size of 2 mm was used as baseline dose distribution. The input included the dose distribution with 4 mm grid size and CT images. The model performance was compared with HD U-Net and cubic interpolation methods using Dose-volume histograms (DVH) metrics and global gamma analysis with 1%/1 mm and 10% low dose threshold. The correlation between the prediction error and the dose, dose gradient, and CT values was also evaluated.

Results

The prediction errors of MDSR were 0.06–0.84% of D_mean indices, and the gamma passing rate was 83.1–91.0% on the benchmark testing dataset, and 0.02–1.03% and 71.3–90.3% for the generalization dataset respectively. The model performance was significantly higher than the HD U-Net and interpolation methods (p < 0.05). The mean errors of the MDSR model decreased (monotonously by 0.03–0.004%) with dose and increased (by 0.01–0.73%) with the dose gradient. There was no correlation between prediction errors and the CT values.

Conclusion

The proposed MDSR model achieved good agreement with the baseline high-resolution dose distribution, with small prediction errors for DVH indices and high gamma passing rate for both seen and unseen sites, indicating a robust and generalizable dose prediction model. The model can provide fast and accurate high-resolution dose distribution for clinical dose calculation, particularly for the routine practice of OLART.

An accurate prognosis of renal function decline in Autosomal Dominant Polycystic Kidney Disease (ADPKD) is crucial for early intervention. Current biomarkers used are height-adjusted total kidney volume (HtTKV), estimated glomerular filtration rate (eGFR), and patient age. However, manually measuring kidney volume is time-consuming and subject to observer variability. Additionally, incorporating automatically generated features from kidney MRI images, along with conventional biomarkers, can enhance prognostic improvement. To address these issues, we developed two deep-learning algorithms. Firstly, an automated kidney volume segmentation model accurately calculates HtTKV. Secondly, we utilize segmented kidney volumes, predicted HtTKV, age, and baseline eGFR to predict chronic kidney disease (CKD) stages $>$=3A, $>$=3B, and a 30% decline in eGFR after 8 years from the baseline visit. Our approach combines a convolutional neural network (CNN) and a multi-layer perceptron (MLP). Our study included 135 subjects and the AUC scores obtained were 0.96, 0.96, and 0.95 for CKD stages $>$=3A, $>$=3B, and a 30% decline in eGFR, respectively. Furthermore, our algorithm achieved a Pearson correlation coefficient of 0.81 between predicted and measured eGFR decline. We extended our approach to predict distinct CKD stages after eight years with an AUC of 0.97. The proposed approach has the potential to enhance monitoring and facilitate prognosis in ADPKD patients, even in the early disease stages.

Introduction

Neuroinflammation evaluation after acute ischemic stroke is a promising option for selecting an appropriate post-stroke treatment strategy. To assess neuroinflammation in vivo, translocator protein PET (TSPO PET) can be used. However, the gold standard TSPO PET quantification method includes a 90 min scan and continuous arterial blood sampling, which is challenging to perform on a routine basis. In this work, we determine what information is required for a simplified quantification approach using a machine learning algorithm.

Materials and Methods

We analyzed data from 18 patients with ischemic stroke who received 0–90 min [¹⁸F]GE-180 PET as well as T1-weigted (T1w), FLAIR, and arterial spin labeling (ASL) MRI scans. During PET scans, five manual venous blood samples at 5, 15, 30, 60, and 85 min post injection (p.i.) were drawn, and plasma activity concentration was measured. Total distribution volume (V_T) was calculated using Logan plot with the full dynamic PET and an image-derived input function (IDIF) from the carotid arteries. IDIF was scaled by a calibration factor derived from all the measured plasma activity concentrations. The calculated V_T values were used for training a random forest regressor. As input features for the model, we used three late PET frames (60–70, 70–80, and 80–90 min p.i.), the ASL image reflecting perfusion, the voxel coordinates, the lesion mask, and the five plasma activity concentrations. The algorithm was validated with the leave-one-out approach. To estimate the impact of the individual features on the algorithm’s performance, we used Shapley Additive Explanations (SHAP). Having determined that the three late PET frames and the plasma activity concentrations were the most important features, we tested a simplified quantification approach consisting of dividing a late PET frame by a plasma activity concentration. All the combinations of frames/samples were compared by means of concordance correlation coefficient and Bland-Altman plots.

Results

When using all the input features, the algorithm predicted V_T values with high accuracy (87.8 ± 8.3%) for both lesion and non-lesion voxels. The SHAP values demonstrated high impact of the late PET frames (60–70, 70–80, and 80–90 min p.i.) and plasma activity concentrations on the V_T prediction, while the influence of the ASL-derived perfusion, voxel coordinates, and the lesion mask was low. Among all the combinations of the late PET frames and plasma activity concentrations, the 70–80 min p.i. frame divided by the 30 min p.i. plasma sample produced the closest V_T estimate in the ischemic lesion.

Conclusion

Reliable TSPO PET quantification is achievable by using a single late PET frame divided by a late blood sample activity concentration.

Multimodal image registration is applied in medical image analysis as it allows the integration of complementary data from multiple imaging modalities. In recent years, various neural network-based approaches for medical image registration have been presented in papers, but due to the use of different datasets, a fair comparison is not possible. In this research 20 different neural networks for an affine registration of medical images were implemented. The networks’ performance and the networks’ generalizability to new datasets were evaluated using two multimodal datasets - a synthetic and a real patient dataset - of three-dimensional CT and MR images of the liver. The networks were first trained semi-supervised using the synthetic dataset and then evaluated on the synthetic dataset and the unseen patient dataset. Afterwards, the networks were finetuned on the patient dataset and subsequently evaluated on the patient dataset. The networks were compared using our own developed CNN as benchmark and a conventional affine registration with SimpleElastix as baseline. Six networks improved the pre-registration Dice coefficient of the synthetic dataset significantly (p-value $<$ 0.05) and nine networks improved the pre-registration Dice coefficient of the patient dataset significantly and are therefore able to generalize to the new datasets used in our experiments. Many different machine learning-based methods have been proposed for affine multimodal medical image registration, but few are generalizable to new data and applications. It is therefore necessary to conduct further research in order to develop medical image registration techniques that can be applied more widely.

Today, as in every life-threatening disease, early diagnosis of brain tumors plays a life-saving role. The brain tumor is formed by the transformation of brain cells from their normal structures into abnormal cell structures. These formed abnormal cells begin to form in masses in the brain regions. Nowadays, many different techniques are employed to detect these tumor masses, and the most common of these techniques is Magnetic Resonance Imaging (MRI). In this study, it is aimed to automatically detect brain tumors with the help of ensemble deep learning architectures (ResNet50, VGG19, InceptionV3 and MobileNet) and Class Activation Maps (CAMs) indicators by employing MRI images. The proposed system was implemented in three stages. In the first stage, it was determined whether there was a tumor in the MR images (Binary Approach). In the second stage, different tumor types (Normal, Glioma Tumor, Meningioma Tumor, Pituitary Tumor) were detected from MR images (Multi-class Approach). In the last stage, CAMs of each tumor group were created as an alternative tool to facilitate the work of specialists in tumor detection. The results showed that the overall accuracy of the binary approach was calculated as 100% on the ResNet50, InceptionV3 and MobileNet architectures, and 99.71% on the VGG19 architecture. Moreover, the accuracy values of 96.45% with ResNet50, 93.40% with VGG19, 85.03% with InceptionV3 and 89.34% with MobileNet architectures were obtained in the multi-class approach.

Purpose

This research aims to develop a feature-guided deep learning approach and compare it with an optimized conventional post-processing algorithm in order to enhance the image quality of diffusion-weighted liver images and, in particular, to reduce the pulsation-induced signal loss occurring predominantly in the left liver lobe.

Methods

Data from 40 patients with liver lesions were used. For the conventional approach, the best-suited out of five examined algorithms was chosen. For the deep learning approach, a U-Net was trained. Instead of learning “gold-standard” target images, the network was trained to optimize four image features (lesion CNR, vessel darkness, data consistency, and pulsation artifact reduction), which could be assessed quantitatively using manually drawn ROIs. A quality score was calculated from these four features. As an additional quality assessment, three radiologists rated different features of the resulting images.

Results

The conventional approach could substantially increase the lesion CNR and reduce the pulsation-induced signal loss. However, the vessel darkness was reduced. The deep learning approach increased the lesion CNR and reduced the signal loss to a slightly lower extent, but it could additionally increase the vessel darkness. According to the image quality score, the quality of the deep-learning images was higher than that of the images obtained using the conventional approach. The radiologist ratings were mostly consistent with the quantitative scores, but the overall quality ratings differed among the readers.

Conclusion

Unlike the conventional algorithm, the deep-learning algorithm increased the vessel darkness. Therefore, it may be a viable alternative to conventional algorithms.

The use of artificial intelligence systems in clinical routine is still hampered by the necessity of a medical device certification and/or by the difficulty of implementing these systems in a clinic’s quality management system. In this context, the key questions for a user are how to ensure robust model predictions and how to appraise the quality of a model’s results on a regular basis.

In this paper we discuss some conceptual foundation for a clinical implementation of a machine learning system and argue that both vendors and users should take certain responsibilities, as is already common practice for high-risk medical equipment.

We propose the methodology from AAPM Task Group 100 report No. 283 as a conceptual framework for developing risk-driven a quality management program for a clinical process that encompasses a machine learning system. This is illustrated with an example of a clinical workflow. Our analysis shows how the risk evaluation in this framework can accommodate artificial intelligence based systems independently of their robustness evaluation or the user’s in–house expertise. In particular, we highlight how the degree of interpretability of a machine learning system can be systematically accounted for within the risk evaluation and in the development of a quality management system.