MRI Segmentation Research Articles

Explainable AI-based Alzheimer's prediction and management using multimodal data.

According to the World Health Organization (WHO), dementia is the seventh leading reason of death among all illnesses and one of the leading causes of disability among the world's elderly people. Day by day the number of Alzheimer's patients is rising. Considering the increasing rate and the dangers, Alzheimer's disease should be diagnosed carefully. Machine learning is a potential technique for Alzheimer's diagnosis but general users do not trust machine learning models due to the black-box nature. Even, some of those models do not provide the best performance because of using only neuroimaging data. To solve these issues, this paper proposes a novel explainable Alzheimer's disease prediction model using a multimodal dataset. This approach performs a data-level fusion using clinical data, MRI segmentation data, and psychological data. However, currently, there is very little understanding of multimodal five-class classification of Alzheimer's disease. For predicting five class classifications, 9 most popular Machine Learning models are used. These models are Random Forest (RF), Logistic Regression (LR), Decision Tree (DT), Multi-Layer Perceptron (MLP), K-Nearest Neighbor (KNN), Gradient Boosting (GB), Adaptive Boosting (AdaB), Support Vector Machine (SVM), and Naive Bayes (NB). Among these models RF has scored the highest value. Besides for explainability, SHapley Additive exPlanation (SHAP) is used in this research work. The performance evaluation demonstrates that the RF classifier has a 10-fold cross-validation accuracy of 98.81% for predicting Alzheimer's disease, cognitively normal, non-Alzheimer's dementia, uncertain dementia, and others. In addition, the study utilized Explainable Artificial Intelligence based on the SHAP model and analyzed the causes of prediction. To the best of our knowledge, we are the first to present this multimodal (Clinical, Psychological, and MRI segmentation data) five-class classification of Alzheimer's disease using Open Access Series of Imaging Studies (OASIS-3) dataset. Besides, a novel Alzheimer's patient management architecture is also proposed in this work.

A regional brain volume-based age prediction model for neonates and the derived brain maturation index.

To develop a postmenstrual age (PMA) prediction model based on segmentation volume and to evaluate the brain maturation index using the proposed model. Neonatal brain MRIs without clinical illness or structural abnormalities were collected from four datasets from the Developing Human Connectome Project, the Catholic University of Korea, Hammersmith Hospital (HS), and Dankook University Hospital (DU). T1- and T2-weighted images were used to train a brain segmentation model. Another model to predict the PMA of neonates based on segmentation data was developed. Accuracy was assessed using mean absolute error (MAE), root mean square error (RMSE), and mean error (ME). The brain maturation index was calculated as the difference between the PMA predicted by the model and the true PMA, and its correlation with postnatal age was analyzed. A total of 247 neonates (mean gestation age 37 ± 4weeks; range 24-42weeks) were included. Thirty-one features were extracted from each neonate and the three most contributing features for PMA prediction were the right lateral ventricle, left caudate, and corpus callosum. The predicted and true PMA were positively correlated (coefficient = 0.88, p < .001). MAE, RMSE, and ME of the external dataset of HS and DU were 1.57 and 1.33, 1.79 and 1.37, and 0.37 and 0.06weeks, respectively. The brain maturation index negatively correlated with postnatal age (coefficient = - 0.24, p < .001). A model that calculates the regional brain volume can predict the PMA of neonates, which can then be utilized to show the brain maturation degree. A brain maturity index based on regional volume of neonate's brain can be used to measure brain maturation degree, which can help identify the status of early brain development. • Neonatal brain MRI segmentation model could be used to assess neonatal brain maturation status. • A postmenstrual age (PMA) prediction model was developed based on a neonatal brain MRI segmentation model. • The brain maturation index, derived from the PMA prediction model, enabled the estimation of the neonatal brain maturation status.

A few-shot learning-based ischemic stroke segmentation system using weighted MRI fusion

Stroke, particularly ischemic stroke, is a major cause of disability and one of the leading causes of adult mortality worldwide. Early and prompt management of stroke patients can reduce the severity of the disease. Doctors usually determine the severity of a stroke by focusing on the region of interest (ROI) in the MRI or CT scan images. An accurate and effective automatic image segmentation system can assist medical professionals as well as automatic detection and classification systems. Deep learning is the current advanced approach for dealing with machine learning and artificial intelligence. However, conventional deep learning requires a large amount of data for training, and the amount of labeled data in the medical field is limited. In this paper, we propose a few-shot learning strategy and integrate it with a base convolutional neural network model, which utilizes a self-attention mechanism to segment MRI for ischemic stroke. By combining the base model with self-attention, we can focus more on the ROI and disregard less important features. Additionally, the proposed system only selects slices with lesions and ignores unlesioned slices. This helps to improve efficiency and reduce the computational load by eliminating the need to tune unnecessary parameters. To achieve even better results, the system also combines two weighted images, FLAIR and DWI, as an early fusion process. Experiments have shown that this approach leads to higher performance compared to using the same system without fusion. The proposed system is evaluated using a publicly available dataset, ISLES 2015 SSIS, and compared with other state-of-the-art (SOTA) systems. It achieves a dice coefficient score of 0.68, which is significantly better than that of other SOTA systems.

The Impact of Fatty Infiltration on MRI Segmentation of Lower Limb Muscles in Neuromuscular Diseases: A Comparative Study of Deep Learning Approaches

The Impact of Fatty Infiltration on MRI Segmentation of Lower Limb Muscles in Neuromuscular Diseases: A Comparative Study of Deep Learning Approaches

Pilot study: is there an influence of lower limb positioning during magnetic resonance imaging on muscle cross section shape assessment in the thigh?

Positioning in an MRI can influence quantitative measures of the muscle. The goal of this pilot study was to assess the influence of different levels of knee elevation during MRI on the predicted cross-sectional muscle shape in the thigh. Data were acquired in three healthy male participants (age: 29.3 ± 5.1y, height: 181.3 ± 6.4cm, weight: 85.1 ± 3.7kg). For each participant, three MRI scans were taken by a trained radiographer with low, moderate and high knee elevation. The shape of the anatomical cross-sectional areas of the hamstrings and quadriceps in three leg positionings were compared by fitting ellipsoidal functions to the segmented MRI data and calculating the so-called J index for every image slice using the Python scripting language. Different levels of knee elevation resulted in apparent changes in J index for all muscles except vastus medialis. Thereby, the changes were overall more pronounced in the hamstrings compared to the quadriceps. Particularly, by elevating the knee from 8 to 15 degree, the percentage changes in J index were between 7.2 and 13.6% for the hamstrings and between 0.5 and 3.3% for the quadriceps, respectively. For assessing the musculoskeletal properties by means of MRI, a standardized positioning of the leg is required and the knee joint angle should be controlled.

Dorsal subthalamic nucleus targeting in deep brain stimulation: microelectrode recording versus 7-Tesla connectivity.

Connectivity-derived 7-Tesla MRI segmentation and intraoperative microelectrode recording can both assist subthalamic nucleus targeting for deep brain stimulation in Parkinson's disease. It remains unclear whether deep brain stimulation electrodes placed in the 7-Tesla MRI segmented subdivision with predominant projections to cortical motor areas (hyperdirect pathway) achieve superior motor improvement and whether microelectrode recording can accurately distinguish the motor subdivision. In 25 patients with Parkinson's disease, deep brain stimulation electrodes were evaluated for being inside or outside the predominantly motor-connected subthalamic nucleus (motor-connected subthalamic nucleus or non-motor-connected subthalamic nucleus, respectively) based on 7-Tesla MRI connectivity segmentation. Hemi-body motor improvement (Movement Disorder Society Unified Parkinson's Disease Rating Scale, Part III) and microelectrode recording characteristics of multi- and single-unit activities were compared between groups. Deep brain stimulation electrodes placed in the motor-connected subthalamic nucleus resulted in higher hemi-body motor improvement, compared with electrodes placed in the non-motor-connected subthalamic nucleus (80% versus 52%, P < 0.0001). Multi-unit activity was found slightly higher in the motor-connected subthalamic nucleus versus the non-motor-connected subthalamic nucleus (P < 0.001, receiver operating characteristic 0.63); single-unit activity did not differ between groups. Deep brain stimulation in the connectivity-derived 7-Tesla MRI subthalamic nucleus motor segment produced a superior clinical outcome; however, microelectrode recording did not accurately distinguish this subdivision within the subthalamic nucleus.

Kernel induced semi-supervised spatial clustering: a novel brain MRI segmentation technique

Kernel induced semi-supervised spatial clustering: a novel brain MRI segmentation technique

Comparison of machine learning models for MRI image-based brain tumor classification and segmentation

Brain tumors have a high-risk factor and are extremely harmful to the human body. With the development of science and technology in recent years, automatic segmentation has become popular in medical diagnosis because it provides higher accuracy than traditional hand segmentation. At present, more and more people start to study and improve it. Due to the non-invasive nature of MRI, MR images are often used to segment and classify brain tumors. However, limited by the inaccuracy and inoperability of manual segmentation, it is very necessary to have a complete and comprehensive automatic brain tumor segmentation and classification algorithm technology. This article discusses the benefits, drawbacks, and areas of application of several traditional algorithms as well as more modern, improved, and more advanced algorithms. Segmentation methods and classification methods can be used to classify these techniques. Convolutional neural networks (CNN), Support vector machines, and Transformers are examples of classification methods. Random forests, decision trees, and improved U-Net algorithms are examples of segmentation methods. To discuss the capability of classification and segmentation, there are three sections in the area used for segmenting brain tumors with three types, including Tumor Core, Enhance Tumor, and Whole Tumor, which could be abbreviated as TC, ET, and WT. Through the comparative analysis of these methods, useful insights for future research are provided.

Automatic Myocardium Segmentation in Delayed-Enhancement MRI with Pathology-Specific Data Augmentation and Deep Learning Architectures

The extent of myocardial infarction (MI) can be evaluated thanks to delayed enhancement (DE) cardiac MRI. DE MRI is an imaging technique acquired several minutes after the injection of a contrast agent where MI appears with a bright signal. The automatic myocardium segmentation in DE MRI is quite challenging, especially when MI is present, since these areas usually showcase a heterogeneous aspect in terms of shape and intensity, thus obstructing the myocardium visibility. To overcome this issue, we propose an image processing-based data augmentation algorithm where diverse synthetic cases of MI were created in two different ways: fixed and adaptive. In the first one, the training set is enlarged by a specific factor, whereas in the second, the method receives feedback from the segmentation model during training and performs the augmentation exclusively on complex cases. The method performance was evaluated in single and multi-modality settings. In this latter, information from kinetic images (Cine MRI), which are acquired along DE MRI in the same examination, is also used, and the extracted features from both modalities are fused. The results show that applying the data augmentation in a fixed fashion on a multi-modality setting leads to a more consistent segmentation of the myocardium in DE MRI. The segmentation models, which were all UNet-based architectures, can better relate MI areas with the myocardium, thus increasing its overall robustness to pathology-specific local pattern perturbations.

Statin use and longitudinal changes in quantitative MRI-based biomarkers of thigh muscle quality: data from Osteoarthritis Initiative.

To assess whetherchanges in MRI-based measures of thigh muscle quality associated with statin use in participants with and without/at-risk of knee osteoarthritis. This retrospective cohort study used data from the Osteoarthritis Initiative study. Statin users and non-users were matched for relevant covariates using 1:1 propensity-score matching. Participants were further stratified according to baseline radiographic knee osteoarthritis status. We used a validated deep-learning method for thigh muscle MRI segmentation and calculation of muscle quality biomarkers at baseline, 2nd, and 4th visits. Mean difference and 95% confidence intervals (CI) in longitudinal 4-year measurements of muscle quality biomarkers, including cross-sectional area, intramuscular adipose tissue, contractile percent, and knee extensors and flexors maximum and specific contractile force (force/muscle area) were the outcomes of interest. After matching, 3772 thighs of 1910 participants were included (1886 thighs of statin-users: 1886 of non-users; age: 62 ± 9 years (average ± standard deviation), range: 45-79; female/male: 1). During 4 years, statin use was associated with a slight decrease in muscle quality, indicated by decreased knee extension maximum (mean-difference, 95% CI: - 1.85 N/year, - 3.23 to - 0.47) and specific contractile force (- 0.04 N/cm2/year, - 0.07 to - 0.01),decreased thigh muscle contractile percent (- 0.03%/year, - 0.06 to - 0.01), and increased thigh intramuscular adipose tissue (3.06 mm2/year, 0.53to 5.59). Stratified analyses showed decreased muscle quality only in participants without/at-risk of knee osteoarthritis but not those with established knee osteoarthritis. Statin use is associated with a slight decrease in MRI-based measures of thigh muscle quality over 4 years. However, considering statins' substantial cardiovascular benefits, these slight muscle changes may be relatively less important in overall patient care.

Radiologically based automated segmentation of cardiac MRI using an improved U-Net neural algorithm

ObjectiveTo overcome the complexity and variability of the heart, along with factors like fuzzy boundaries and low contrast and produce automated segmentation based on machine learning. MethodsSegmentation of cardiac MRI plays a vital role in various clinical applications. In this study, we propose an improved U-Net-CSP (Cross Stage Partial) method for accurate and robust segmentation of cardiac MRI images.The U-Net-CSP architecture combines the U-Net framework with the CSP module to enhance the segmentation performance. The CSP module enables improved feature reuse and mitigates overfitting issues associated with deep networks. By incorporating the CSP module in both the encoding and decoding stages, the U-Net-CSP effectively captures complex features and preserves important details during the upsampling and downsampling processes.To optimize the segmentation results, we utilize a combination of weighted cross-entropy loss, dice loss, and L2 regularization as the loss function. This loss function facilitates accurate delineation of the regions of interest and enhances the segmentation performance of the U-Net-CSP model. ResultsThe proposed U-Net-CSP method achieves superior performance in cardiac MRI segmentation, surpassing conventional U-Net models and state-of-the-art methods. It demonstrates potential for clinical applications in cardiac therapy, disease research, and model reconstruction. Our method automates segmentation of the left ventricle, right ventricle, and aorta from cardiac MRI images. Comparative analyses with leading methods show consistent favorable outcomes under diverse conditions. The U-Net-CSP approach accurately delineates cardiac structures from MRI scans, contributing to medical image analysis and advancing cardiac healthcare. ConclusionIn conclusion, the proposed U-Net-CSP method provides an enhanced approach for cardiac MRI segmentation, overcoming the challenges posed by complex cardiac structures and image characteristics. This study contributes to the advancement of medical image analysis techniques, particularly in the field of cardiac MRI segmentation.

The utility of automatic segmentation of kidney MRI in chronic kidney disease using a 3D convolutional neural network

We developed a 3D convolutional neural network (CNN)-based automatic kidney segmentation method for patients with chronic kidney disease (CKD) using MRI Dixon-based T1-weighted in-phase (IP)/opposed-phase (OP)/water-only (WO) images. The dataset comprised 100 participants with renal dysfunction (RD; eGFR < 45 mL/min/1.73 m2) and 70 without (non-RD; eGFR ≥ 45 mL/min/1.73 m2). The model was applied to the right, left, and both kidneys; it was first evaluated on the non-RD group data and subsequently on the combined data of the RD and non-RD groups. For bilateral kidney segmentation of the non-RD group, the best performance was obtained when using IP image, with a Dice score of 0.902 ± 0.034, average surface distance of 1.46 ± 0.75 mm, and a difference of − 27 ± 21 mL between ground-truth and automatically computed volume. Slightly worse results were obtained for the combined data of the RD and non-RD groups and for unilateral kidney segmentation, particularly when segmenting the right kidney from the OP images. Our 3D CNN-assisted automatic segmentation tools can be utilized in future studies on total kidney volume measurements and various image analyses of a large number of patients with CKD.

A Symmetrical Approach to Brain Tumor Segmentation in MRI Using Deep Learning and Threefold Attention Mechanism

The symmetrical segmentation of brain tumor images is crucial for both clinical diagnosis and computer-aided prognosis. Traditional manual methods are not only asymmetrical in terms of efficiency but also prone to errors and lengthy processing. A significant barrier to the process is the complex interplay between the deep learning network for MRI brain tumor imaging and the harmonious compound of both local and global feature information, which can throw off the balance in segmentation accuracy. Addressing this asymmetry becomes essential for precise diagnosis. In answer to this challenge, we introduce a balanced, end-to-end solution for brain tumor segmentation, incorporating modifications that mirror the U-Net architecture, ensuring a harmonious flow of information. Beginning with symmetric enhancement of the visual quality of MRI brain images, we then apply a symmetrical residual structure. By replacing the convolutional modules in both the encoder and decoder sections with deep residual modules, we establish a balance that counters the vanishing gradient problem commonly faced when the network depth increases. Following this, a symmetrical threefold attention block is integrated. This addition ensures a balanced fusion of local and global image features, fine-tuning the network to symmetrically discern and learn essential image characteristics. This harmonious integration remarkably amplifies the network’s precision in segmenting MRI brain tumors. We further validate the equilibrium achieved by our proposed model using three brain tumor segmentation datasets and four metrics and by juxtaposing our model against 21 traditional and learning-based counterparts. The results confirm that our balanced approach significantly elevates performance in the segmentation of MRI brain tumor images without an asymmetrical increase in computational time.

Stochastic co-teaching for training neural networks with unknown levels of label noise

Label noise hampers supervised training of neural networks. However, data without label noise is often infeasible to attain, especially for medical tasks. Attaining high-quality medical labels would require a pool of experts and their consensus reading, which would be extremely costly. Several methods have been proposed to mitigate the adverse effects of label noise during training. State-of-the-art methods use multiple networks that exploit different decision boundaries to identify label noise. Among the best performing methods is co-teaching. However, co-teaching comes with the requirement of knowing label noise a priori. Hence, we propose a co-teaching method that does not require any prior knowledge about the level of label noise. We introduce stochasticity to select or reject training instances. We have extensively evaluated the method on synthetic experiments with extreme label noise levels and applied it to real-world medical problems of ECG classification and cardiac MRI segmentation. Results show that the approach is robust to its hyperparameter choice and applies to various classification tasks with unknown levels of label noise.

Syn_SegNet: A Joint Deep Neural Network for Ultrahigh-Field 7T MRI Synthesis and Hippocampal Subfield Segmentation in Routine 3T MRI.

Precise delineation of hippocampus subfields is crucial for the identification and management of various neurological and psychiatric disorders. However, segmenting these subfields automatically in routine 3T MRI is challenging due to their complex morphology and small size, as well as the limited signal contrast and resolution of the 3T images. This research proposes Syn_SegNet, an end-to-end, multitask joint deep neural network that leverages ultrahigh-field 7T MRI synthesis to improve hippocampal subfield segmentation in 3T MRI. Our approach involves two key components. First, we employ a modified Pix2PixGAN as the synthesis model, incorporating self-attention modules, image and feature matching loss, and ROI loss to generate high-quality 7T-like MRI around the hippocampal region. Second, we utilize a variant of 3D-U-Net with multiscale deep supervision as the segmentation subnetwork, incorporating an anatomic weighted cross-entropy loss that capitalizes on prior anatomical knowledge. We evaluate our method on hippocampal subfield segmentation in paired 3T MRI and 7T MRI with seven different anatomical structures. The experimental findings demonstrate that Syn_SegNet's segmentation performance benefits from integrating synthetic 7T data in an online manner and is superior to competing methods. Furthermore, we assess the generalizability of the proposed approach using a publicly accessible 3T MRI dataset. The developed method would be an efficient tool for segmenting hippocampal subfields in routine clinical 3T MRI.

Federated Semi-Supervised Learning for Medical Image Segmentation via Pseudo-Label Denoising.

Distributed big data and digital healthcare technologies have great potential to promote medical services, but challenges arise when it comes to learning predictive model from diverse and complex e-health datasets. Federated Learning (FL), as a collaborative machine learning technique, aims to address the challenges by learning a joint predictive model across multi-site clients, especially for distributed medical institutions or hospitals. However, most existing FL methods assume that clients possess fully labeled data for training, which is often not the case in e-health datasets due to high labeling costs or expertise requirement. Therefore, this work proposes a novel and feasible approach to learn a Federated Semi-Supervised Learning (FSSL) model from distributed medical image domains, where a federated pseudo-labeling strategy for unlabeled clients is developed based on the embedded knowledge learned from labeled clients. This greatly mitigates the annotation deficiency at unlabeled clients and leads to a cost-effective and efficient medical image analysis tool. We demonstrated the effectiveness of our method by achieving significant improvements compared to the state-of-the-art in both fundus image and prostate MRI segmentation tasks, resulting in the highest Dice scores of 89.23% and 91.95% respectively even with only a few labeled clients participating in model training. This reveals the superiority of our method for practical deployment, ultimately facilitating the wider use of FL in healthcare and leading to better patient outcomes.

Adaptive Auto-Segmentation for MRI-Guided Online Adaptive Radiotherapy of Cervical Carcinoma

Accurate and efficient delineation of organs and targets on session images is critical in MRI-guided online adaptive radiotherapy (MRgOART). This study proposes a registration-guided deep learning image segmentation framework to assist online delineation of cervical carcinoma. A total of 300 T2-weighted MR images were acquired for patients with cervical carcinoma treated by a 1.5T Unity MR-Linac. The CTV, bladder, rectum, pelvic bone and femoral joints were delineated on each MRI by the same radiation oncologist. To overcome these obstacles to online MRI segmentation, we propose a registration-guided DL (RgDL) segmentation framework that integrates image registration algorithms and DL segmentation models. Firstly, the DL segmentation model was trained using nnU-net. Then, for each treatment fraction, the deformable image registration (DIR) algorithm generates initial contours from previous treatment fraction, which were used as guidance by DL model to obtain the accurate current segmentation. The segmentation accuracy of alone DIR, DL and RgDL were evaluated by dice similarity coefficients (DSC) and other distance-based metrics. Compared to the baseline approaches using the DIR and the DL alone, RgDL achieved a DSC of 91.12% on CTV, higher than DIR and DL alone by 15.54% and 10.13%. The DSC of RgDL were improved to 95.58%, 93.65%, 87.8% and 94.84% for bladder, pelvic bone, rectum and femoral joints, higher than DIR and DL alone by 9.61% on average. The proposed adaptive auto-segmentation method can achieve accurate and efficient segmentation and potentially overcome these obstacles to MRgOART.

A novel approach for brain MRI segmentation and image restoration under intensity inhomogeneity and noisy conditions

Brain magnetic resonance (MR) images are vulnerable to many photometric artifacts such as noise, and intensity inhomogeneity (IH). Furthermore, they contain brain-related intrinsic structural complexities such as edges, corners, and other fine structures. These artifacts and structural complexities pose several challenges for the segmentation process of brain MR images which is needed for the proper diagnosis of brain related diseases. We develop a segmentation framework which deals with noise, IH, uncertainty, and other structural complexities present in the brain MR images in unison. The noise is dealt with by incorporating the local spatial and gray level information in the form of a local parameter-free fuzzy factor. This factor also preserves the fine structural details that are present in abundance in the brain MR images. A recently introduced approach for dealing with the IH is incorporated, and integrated with the local information. The problem pertaining to uncertainty in assigning membership to the pixels lying near the boundaries of different tissues is solved by representing the various components of the information such as intensity, cluster centers, and IH, using the intuitionistic fuzzy set theory. Further, a procedure is developed to obtain artifact-free image which could be useful for its comparison with the input image for the better analysis of the MR images by the experts. Thus, an integrated approach is presented to segment and restore the MR images. Experimental results demonstrate the superiority of the proposed approach over the state-of-the-art approaches about segmentation accuracy and robustness to the various imaging artifacts.

Advancements in hybrid approaches for brain tumor segmentation in MRI: a comprehensive review of machine learning and deep learning techniques

Advancements in hybrid approaches for brain tumor segmentation in MRI: a comprehensive review of machine learning and deep learning techniques

Deep learning-based segmentation of whole-body fetal MRI and fetal weight estimation: assessing performance, repeatability, and reproducibility.

To develop a deep-learning method for whole-body fetal segmentation based on MRI; to assess the method's repeatability, reproducibility, and accuracy; to create an MRI-based normal fetal weight growth chart; and to assess the sensitivity to detect fetuses with growth restriction (FGR). Retrospective data of 348 fetuses with gestational age (GA) of 19-39weeks were included: 249 normal appropriate for GA (AGA), 19 FGR, and 80 Other (having various imaging abnormalities). A fetal whole-body segmentation model with a quality estimation module was developed and evaluated in 169 cases. The method was evaluated for its repeatability (repeated scans within the same scanner, n = 22), reproducibility (different scanners, n = 6), and accuracy (compared with birth weight, n = 7). A normal MRI-based growth chart was derived. The method achieved a Dice = 0.973, absolute volume difference ratio (VDR) = 1.8% and VDR mean difference = 0.75% ([Formula: see text]: - 3.95%, 5.46), and high agreement with the gold standard. The method achieved a repeatability coefficient = 4.01%, ICC = 0.99, high reproducibility with a mean difference = 2.21% ([Formula: see text]: - 1.92%, 6.35%), and high accuracy with a mean difference between estimated fetal weight (EFW) and birth weight of - 0.39% ([Formula: see text]: - 8.23%, 7.45%). A normal growth chart (n = 246) was consistent with four ultrasound charts. EFW based on MRI correctly predicted birth-weight percentiles for all 18 fetuses ≤ 10thpercentile and for 14 out of 17 FGR fetuses below the 3rd percentile. Six fetuses referred to MRI as AGA were found to be < 3rd percentile. The proposed method for automatic MRI-based EFW demonstrated high performance and sensitivity to identify FGR fetuses. Results from this study support the use of the automatic fetal weight estimation method based on MRI for the assessment of fetal development and to detect fetuses at risk for growth restriction. • An AI-based segmentation method with a quality assessment module for fetal weight estimation based on MRI was developed, achieving high repeatability, reproducibility, and accuracy. • An MRI-based fetal weight growth chart constructed from a large cohort of normal and appropriate gestational-age fetuses is proposed. • The method showed a high sensitivity for the diagnosis of small fetuses suspected of growth restriction.