Generalized likelihood ratio tests for change detection in diffusion tensor images: application to multiple sclerosis

The automatic analysis of subtle changes between MRI scans is an important tool for monitoring disease evolution. Several methods have been proposed to detect changes in serial conventional MRI but few works have considered Diffusion Tensor Imaging (DTI), which is a promising modality for monitoring neurodegenerative disease and particularly Multiple Sclerosis (MS). In this paper, we introduce a comprehensive framework for detecting changes between two DTI acquisitions by considering different levels of representation of diffusion imaging, namely the Apparent Diffusion Coefficient (ADC) images, the diffusion tensor fields, and scalar images characterizing diffusion properties such as the fractional anisotropy and the mean diffusivity. The proposed statistical method for change detection is based on the Generalized Likelihood Ratio Test (GLRT) that has been derived for the different diffusion imaging representations, based on the core assumption of a Gaussian diffusion model and of an additive Gaussian noise on the ADCs. Results on synthetic and real images demonstrate the ability of the different tests to bring useful and complementary information in the context of the follow-up of MS patients.     

Longitudinal change detection in diffusion MRI using multivariate statistical testing on tensors

This paper presents a longitudinal change detection framework for detecting relevant modifications in diffusion MRI, with application to neuromyelitis optica (NMO) and multiple sclerosis (MS). The core problem is to identify image regions that are significantly different between two scans. The proposed method is based on multivariate statistical testing which was initially introduced for tensor population comparison. We use this method in the context of longitudinal change detection by considering several strategies to build sets of tensors characterizing the variability of each voxel. These strategies make use of the variability existing in the diffusion weighted images (thanks to a bootstrap procedure), or in the spatial neighborhood of the considered voxel, or a combination of both. Results on synthetic evolutions and on real data are presented. Interestingly, experiments on NMO patients highlight the ability of the proposed approach to detect changes in the normal-appearing white matter (according to conventional MRI) that are related with physical status outcome. Experiments on MS patients highlight the ability of the proposed approach to detect changes in evolving and non-evolving lesions (according to conventional MRI). These findings might open promising prospects for the follow-up of NMO and MS pathologies.     

Exploring the discrimination power of the time domain for segmentation and characterization of active lesions in serial MR data

This paper presents a new method for the automatic segmentation and characterization of object changes in time series of three-dimensional data sets. The technique was inspired by procedures developed for analysis of functional MRI data sets. After precise registration of serial volume data sets to 4-D data, we applied a time series analysis taking into account the characteristic time function of variable lesions. The images were preprocessed with a correction of image field inhomogeneities and a normalization of the brightness over the whole time series. Thus, static regions remain unchanged over time, whereas changes in tissue characteristics produce typical intensity variations in the voxel's time series. A set of features was derived from the time series, expressing probabilities for membership to the sought structures. These multiple sources of uncertain evidence were combined to a single evidence value using Dempster-Shafer's theory. The project was driven by the objective of improving the segmentation and characterization of white matter lesions in serial MR data of multiple sclerosis patients. Pharmaceutical research and patient follow-up requires efficient and robust methods with a high degree of automation. The new approach replaces conventional segmentation of series of 3-D data sets by a 1-D processing of the temporal change at each voxel in the 4-D image data set. The new method has been applied to a total of 11 time series from different patient studies, covering time resolutions of 12 and 24 data sets over a period of about 1 year. The results demonstrate that time evolution is a highly sensitive feature for detection of fluctuating structures.     

SPASM: a 3D-ASM for segmentation of sparse and arbitrarily oriented cardiac MRI data

A new technique (SPASM) based on a 3D-ASM is presented for automatic segmentation of cardiac MRI image data sets consisting of multiple planes with arbitrary orientations, and with large undersampled regions. Model landmark positions are updated in a two-stage iterative process. First, landmark positions close to intersections with images are updated. Second, the update information is propagated to the regions without image information, such that new locations for the whole set of the model landmarks are obtained. Feature point detection is performed by a fuzzy inference system, based on fuzzy C-means clustering. Model parameters were optimized on a computer cluster and the computational load distributed by grid computing. SPASM was applied to image data sets with an increasing sparsity (from 2 to 11 slices) comprising images with different orientations and stemming from different MRI acquisition protocols. Segmentation outcomes and calculated volumes were compared to manual segmentation on a dense short-axis data configuration in a 3D manner. For all data configurations, (sub-)pixel accuracy was achieved. Performance differences between data configurations were significantly different (p<0.05) for SA data sets with less than 6 slices, but not clinically relevant (volume differences<4 ml). Comparison to results from other 3D model-based methods showed that SPASM performs comparable to or better than these other methods, but SPASM uses considerably less image data. Sensitivity to initial model placement proved to be limited within a range of position perturbations of approximately 20 mm in all directions.     

Fully automatic segmentation of the mitral leaflets in 3D transesophageal echocardiographic images using multi-atlas joint label fusion and deformable medial modeling

Comprehensive visual and quantitative analysis of in vivo human mitral valve morphology is central to the diagnosis and surgical treatment of mitral valve disease. Real-time 3D transesophageal echocardiography (3D TEE) is a practical, highly informative imaging modality for examining the mitral valve in a clinical setting. To facilitate visual and quantitative 3D TEE image analysis, we describe a fully automated method for segmenting the mitral leaflets in 3D TEE image data. The algorithm integrates complementary probabilistic segmentation and shape modeling techniques (multi-atlas joint label fusion and deformable modeling with continuous medial representation) to automatically generate 3D geometric models of the mitral leaflets from 3D TEE image data. These models are unique in that they establish a shape-based coordinate system on the valves of different subjects and represent the leaflets volumetrically, as structures with locally varying thickness. In this work, expert image analysis is the gold standard for evaluating automatic segmentation. Without any user interaction, we demonstrate that the automatic segmentation method accurately captures patient-specific leaflet geometry at both systole and diastole in 3D TEE data acquired from a mixed population of subjects with normal valve morphology and mitral valve disease.     

CEREBRUM: a fast and fully-volumetric Convolutional Encoder-decodeR for weakly-supervised sEgmentation of BRain strUctures from out-of-the-scanner MRI

Many functional and structural neuroimaging studies call for accurate morphometric segmentation of different brain structures starting from image intensity values of MRI scans. Current automatic (multi-) atlas-based segmentation strategies often lack accuracy on difficult-to-segment brain structures and, since these methods rely on atlas-to-scan alignment, they may take long processing times. Alternatively, recent methods deploying solutions based on Convolutional Neural Networks (CNNs) are enabling the direct analysis of out-of-the-scanner data. However, current CNN-based solutions partition the test volume into 2D or 3D patches, which are processed independently. This process entails a loss of global contextual information, thereby negatively impacting the segmentation accuracy. In this work, we design and test an optimised end-to-end CNN architecture that makes the exploitation of global spatial information computationally tractable, allowing to process a whole MRI volume at once. We adopt a weakly supervised learning strategy by exploiting a large dataset composed of 947 out-of-the-scanner (3 Tesla T1-weighted 1mm isotropic MP-RAGE 3D sequences) MR Images. The resulting model is able to produce accurate multi-structure segmentation results in only a few seconds. Different quantitative measures demonstrate an improved accuracy of our solution when compared to state-of-the-art techniques. Moreover, through a randomised survey involving expert neuroscientists, we show that subjective judgements favour our solution with respect to widely adopted atlas-based software.     

Machine learning and radiology

In this paper, we give a short introduction to machine learning and survey its applications in radiology. We focused on six categories of applications in radiology: medical image segmentation, registration, computer aided detection and diagnosis, brain function or activity analysis and neurological disease diagnosis from fMR images, content-based image retrieval systems for CT or MRI images, and text analysis of radiology reports using natural language processing (NLP) and natural language understanding (NLU). This survey shows that machine learning plays a key role in many radiology applications. Machine learning identifies complex patterns automatically and helps radiologists make intelligent decisions on radiology data such as conventional radiographs, CT, MRI, and PET images and radiology reports. In many applications, the performance of machine learning-based automatic detection and diagnosis systems has shown to be comparable to that of a well-trained and experienced radiologist. Technology development in machine learning and radiology will benefit from each other in the long run. Key contributions and common characteristics of machine learning techniques in radiology are discussed. We also discuss the problem of translating machine learning applications to the radiology clinical setting, including advantages and potential barriers.     

Multi-contrast human neonatal brain atlas: application to normal neonate development analysis

MRI is a sensitive method for detecting subtle anatomic abnormalities in the neonatal brain. To optimize the usefulness for neonatal and pediatric care, systematic research, based on quantitative image analysis and functional correlation, is required. Normalization-based image analysis is one of the most effective methods for image quantification and statistical comparison. However, the application of this methodology to neonatal brain MRI scans is rare. Some of the difficulties are the rapid changes in T1 and T2 contrasts and the lack of contrast between brain structures, which prohibits accurate cross-subject image registration. Diffusion tensor imaging (DTI), which provides rich and quantitative anatomical contrast in neonate brains, is an ideal technology for normalization-based neonatal brain analysis. In this paper, we report the development of neonatal brain atlases with detailed anatomic information derived from DTI and co-registered anatomical MRI. Combined with a diffeomorphic transformation, we were able to normalize neonatal brain images to the atlas space and three-dimensionally parcellate images into 122 regions. The accuracy of the normalization was comparable to the reliability of human raters. This method was then applied to babies of 37-53 post-conceptional weeks to characterize developmental changes of the white matter, which indicated a posterior-to-anterior and a central-to-peripheral direction of maturation. We expect that future applications of this atlas will include investigations of the effect of prenatal events and the effects of preterm birth or low birth weights, as well as clinical applications, such as determining imaging biomarkers for various neurological disorders.     

Template images for nonhuman primate neuroimaging: 2. Macaque

Neuroimaging studies are increasingly performed in macaque species, including the pig-tailed macaque (Macaca nemestrina). At times experimental questions can be answered by analysis of functional images in individual subjects and reference to a structural image in that subject. However, coregistration of functional brain images across many subjects offers the experimental advantage of enabling voxel-based analysis over multiple subjects and is therefore widely used in human studies. Voxel-based coregistration methods require a high-quality 3D template image. We created such templates, derived from T1-weighted MRI and blood-flow PET images from 12 nemestrina monkeys. We designed the macaque templates to be maximally compatible with the baboon template images described in a companion paper, to facilitate cross-species comparison of functional imaging data. Here we present data showing the reliability and validity of automatic image registration to the template. Alignment of selected internal fiducial points was accurate to within 1.9 mm overall (mean) even across species. The template images, along with copies aligned to the UCLA nemestrina brain atlas, are available on the Internet (purl.org/net/kbmd/n2k) and can be used as targets with any image registration software.     

3D Registration of pre-surgical prostate MRI and histopathology images via super-resolution volume reconstruction

The use of MRI for prostate cancer diagnosis and treatment is increasing rapidly. However, identifying the presence and extent of cancer on MRI remains challenging, leading to high variability in detection even among expert radiologists. Improvement in cancer detection on MRI is essential to reducing this variability and maximizing the clinical utility of MRI. To date, such improvement has been limited by the lack of accurately labeled MRI datasets. Data from patients who underwent radical prostatectomy enables the spatial alignment of digitized histopathology images of the resected prostate with corresponding pre-surgical MRI. This alignment facilitates the delineation of detailed cancer labels on MRI via the projection of cancer from histopathology images onto MRI. We introduce a framework that performs 3D registration of whole-mount histopathology images to pre-surgical MRI in three steps. First, we developed a novel multi-image super-resolution generative adversarial network (miSRGAN), which learns information useful for 3D registration by producing a reconstructed 3D MRI. Second, we trained the network to learn information between histopathology slices to facilitate the application of 3D registration methods. Third, we registered the reconstructed 3D histopathology volumes to the reconstructed 3D MRI, mapping the extent of cancer from histopathology images onto MRI without the need for slice-to-slice correspondence. When compared to interpolation methods, our super-resolution reconstruction resulted in the highest PSNR relative to clinical 3D MRI (32.15 dB vs 30.16 dB for BSpline interpolation). Moreover, the registration of 3D volumes reconstructed via super-resolution for both MRI and histopathology images showed the best alignment of cancer regions when compared to (1) the state-of-the-art RAPSODI approach, (2) volumes that were not reconstructed, or (3) volumes that were reconstructed using nearest neighbor, linear, or BSpline interpolations. The improved 3D alignment of histopathology images and MRI facilitates the projection of accurate cancer labels on MRI, allowing for the development of improved MRI interpretation schemes and machine learning models to automatically detect cancer on MRI.     

Dense feature deformation morphometry: Incorporating DTI data into conventional MRI morphometry

Registration based mapping of geometric differences in MRI anatomy allows the detection of subtle and complex changes in brain anatomy over time that provides an important quantitative window on the process of both brain development and degeneration. However, methods developed for this have so far been aimed at using conventional structural MRI data (T1W imaging) and the resulting maps are limited in their ability to localize patterns of change within sub-regions of uniform tissue. Alternative MRI contrast mechanisms, in particular Diffusion Tensor Imaging (DTI) data are now more commonly being used in serial studies and provide valuable complementary microstructural information within white matter. This paper describes a new approach which incorporates information from DTI data into deformation tensor morphometry of conventional MRI. The key problem of using the additional information provided by DTI data is addressed by proposing a novel mutual information (MI) derived criterion termed diffusion paired MI. This combines conventional and diffusion data in a single registration measure. We compare different formulations of this measure when used in a diffeomorphic fluid registration scheme to map local volume changes. Results on synthetic data and example images from clinical studies of neurodegenerative conditions illustrate the improved localization of tissue volume changes provided by the incorporation of DTI data into the morphometric registration.     

Ultrasound virtual endoscopic imaging

Volume data acquisition, three dimensional (3D) imaging, and multiplanar reformatting have become widely used for computed tomography (CT) and magnetic resonance imaging (MRI). As an extension of this technology, virtual endoscopic visualization of hollow organs has become a reality that is now finding its way into clinical CT practice. The same methods of computer processing as are used for CT and MRI can be applied to an ultrasound (US) volume image data set with the same potential output; namely, 3D, multiplanar, and virtual endoscopic images. The use of this image processing technology for US applications has lagged behind the CT and MRI applications, but considerable progress in applying these methods to US has occurred in recent years. As a result, US virtual endoscopic imaging now can be performed on a clinical basis by using standard US instruments and commercially available computer software. The use of newer US imaging methods, such as tissue harmonic and power Doppler imaging, has enhanced the potential for US virtual endoscopy. This article reviews the technology of US virtual endoscopy. In addition, our preliminary experience of using this method for abdominal and vascular diagnosis is described. Finally, we speculate on technical improvements and potential applications that are likely in the future.     

A fully automatic and robust brain MRI tissue classification method

A novel, fully automatic, adaptive, robust procedure for brain tissue classification from 3D magnetic resonance head images (MRI) is described in this paper. The procedure is adaptive in that it customizes a training set, by using a 'pruning' strategy, such that the classification is robust against anatomical variability and pathology. Starting from a set of samples generated from prior tissue probability maps (a 'model') in a standard, brain-based coordinate system ('stereotaxic space'), the method first reduces the fraction of incorrectly labeled samples in this set by using a minimum spanning tree graph-theoretic approach. Then, the corrected set of samples is used by a supervised kNN classifier for classifying the entire 3D image. The classification procedure is robust against variability in the image quality through a non-parametric implementation: no assumptions are made about the tissue intensity distributions. The performance of this brain tissue classification procedure is demonstrated through quantitative and qualitative validation experiments on both simulated MRI data (10 subjects) and real MRI data (43 subjects). A significant improvement in output quality was observed on subjects who exhibit morphological deviations from the model due to aging and pathology.     

A Review On digital image processing techniques for in-Vivo confocal images of the cornea

This work reviews the scientific literature regarding digital image processing for in vivo confocal microscopy images of the cornea. We present and discuss a selection of prominent techniques designed for semi- and automatic analysis of four areas of the cornea (epithelium, sub-basal nerve plexus, stroma and endothelium). The main context is image enhancement, detection of structures of interest, and quantification of clinical information. We have found that the preprocessing stage lacks of quantitative studies regarding the quality of the enhanced image, or its effects in subsequent steps of the image processing. Threshold values are widely used in the reviewed methods, although generally, they are selected empirically and manually. The image processing results are evaluated in many cases through comparison with gold standards not widely accepted. It is necessary to standardize values to be quantified in terms of sensitivity and specificity of methods. Most of the reviewed studies do not show an estimation of the computational cost of the image processing. We conclude that reliable, automatic, computer-assisted image analysis of the cornea is still an open issue, constituting an interesting and worthwhile area of research.     

Automated MR image classification in temporal lobe epilepsy

In those with drug refractory focal epilepsy, MR imaging is important for identifying structural causes of seizures that may be amenable to surgical treatment. In up to 25% of potential surgical candidates, however, MRI is reported as unremarkable even when employing epilepsy specific sequences. Automated MRI classification is a desirable tool to augment the interpretation of images, especially when changes are subtle or distributed and may be missed on visual inspection. Support vector machines (SVM) have recently been described to be useful for voxel-based MR image classification. In the present study we sought to evaluate whether this method is feasible in temporal lobe epilepsy, with adequate accuracy.We studied 38 patients with hippocampal sclerosis and unilateral (mesial) temporal lobe epilepsy (mTLE) (20 left) undergoing presurgical evaluation and 22 neurologically normal control subjects. 3D T1-weighted images were acquired at 3T (GE Excite), segmented into tissue classes, normalized and smoothed with SPM8. Diffusion tensor imaging (DTI) and double echo images for T2 relaxometry were also acquired and processed. The SVM analysis was done with the libsvm software package in a leave-one-out cross-validation design and predictive accuracy was measured. Local weighting was applied by SPM F-contrast maps.Best accuracies were achieved using the gray matter based segmentation (90-100%) and mean diffusivity (95-97%). For the three-way classification, accuracies were 88 and 93% respectively. Local weighting generally improved the accuracies except in the FA-based processing for which no effect was noted. Removing the hippocampus from the analysis, on the other hand, reduced the obtainable diagnostic indices but these were still > 90% for DTI-based methods and lateralization based on gray matter maps. These findings show that automated SVM image classification can achieve high diagnostic accuracy in mTLE and that voxel-based MRI can be used at the individual subject level. This could be helpful for screening assessments of MRI scans in patients with epilepsy and when no lesion is detected on visual evaluation.     

Automatic curvilinear reformatting of three-dimensional MRI data of the cerebral cortex

Curvilinear reformatting of three-dimensional (3D) MRI data of the cerebral cortex is a well-established tool which improves the display of the gyral structure, permits a precise localization of lesions, and helps to identify subtle abnormalities difficult to detect in planar slices due to the brain's complex convolutional pattern. However, the method is time consuming because it requires interactive manual delineation of the brain surface contour. Therefore, a novel technique for automatic curvilinear reformatting is presented. A T1-weighted MRI volume data set is normalized using SPM2. Due to the normalization to a common stereotactic space, predefined masks can be applied to cover skull and outer brain regions in different depths from the brain surface. Thereby, the outer brain regions are subsequently removed in 2-mm layers parallel to the brain surface like 'peeling an onion'. The serial convex planes enclosing the residual inner part of the brain are presented 3-dimensionally. If necessary (e.g., for intraoperative navigation), the normalized data can be transferred to native space by inverse normalization. Compared to cross-sectional images, curvilinear reformatting offers a markedly superior visualization of topographic relations between lesions and cortical structures, helps to detect subtle cortical malformations and to assess the spatial extent of lesions, thus allowing a better planning of neurosurgical procedures. Compared to alternative methods, it is largely based on freely available software and does not require observer-dependent manual input. In conclusion, we present a simple, easy-to-use and fully automated method for curvilinear reformatting of 3D MRI.     

A tissue-specific adaptive texture filter for medical ultrasound images

The interpretation of ultrasound images remains a difficult task and the opinion of different doctors is generally not unequivocal. Therefore, there is a growing interest in the field of computer-aided diagnosis. In the field of medical image processing, computer-aided diagnosis includes image enhancement to facilitate visual interpretation, automatic indication of affected areas, organs and other regions of medical interest, the performance of automatic measurements and image registration. In this article, we introduce a new algorithm for ultrasound image enhancement that employs a multivariate texture classifier based on the co-occurrence matrix, which, in combination with an adaptive texture smoothing filter, is used to enhance the visual difference between and improve boundary detection between healthy neonatal brain tissue and tissue affected by periventricular leukomalacia. For a quantitative comparison, we delineate the periventricular leukomalacia-affected regions with two different active contours before and after processing 10 images with the proposed technique and several speckle filters from the literature. The semi-automatic delineations thus obtained are compared with the manual delineations of a neonatologist. In all cases, the average delineation achieved with the proposed technique is closer to that of the manual expert delineation than when the images are processed with the other techniques.     

Coronary artery centerline extraction in cardiac CT angiography using a CNN-based orientation classifier

Coronary artery centerline extraction in cardiac CT angiography (CCTA) images is a prerequisite for evaluation of stenoses and atherosclerotic plaque. In this work, we propose an algorithm that extracts coronary artery centerlines in CCTA using a convolutional neural network (CNN).In the proposed method, a 3D dilated CNN is trained to predict the most likely direction and radius of an artery at any given point in a CCTA image based on a local image patch. Starting from a single seed point placed manually or automatically anywhere in a coronary artery, a tracker follows the vessel centerline in two directions using the predictions of the CNN. Tracking is terminated when no direction can be identified with high certainty. The CNN is trained using manually annotated centerlines in training images. No image preprocessing is required, so that the process is guided solely by the local image values around the tracker’s location.The CNN was trained using a training set consisting of 8 CCTA images with a total of 32 manually annotated centerlines provided in the MICCAI 2008 Coronary Artery Tracking Challenge (CAT08). Evaluation was performed within the CAT08 challenge using a test set consisting of 24 CCTA test images in which 96 centerlines were extracted. The extracted centerlines had an average overlap of 93.7% with manually annotated reference centerlines. Extracted centerline points were highly accurate, with an average distance of 0.21 mm to reference centerline points. Based on these results the method ranks third among 25 publicly evaluated methods in CAT08. In a second test set consisting of 50 CCTA scans acquired at our institution (UMCU), an expert placed 5448 markers in the coronary arteries, along with radius measurements. Each marker was used as a seed point to extract a single centerline, which was compared to the other markers placed by the expert. This showed strong correspondence between extracted centerlines and manually placed markers. In a third test set containing 36 CCTA scans from the MICCAI 2014 Challenge on Automatic Coronary Calcium Scoring (orCaScore), fully automatic seeding and centerline extraction was evaluated using a segment-wise analysis. This showed that the algorithm is able to fully-automatically extract on average 92% of clinically relevant coronary artery segments. Finally, the limits of agreement between reference and automatic artery radius measurements were found to be below the size of one voxel in both the CAT08 dataset and the UMCU dataset. Extraction of a centerline based on a single seed point required on average 0.4 ± 0.1 s and fully automatic coronary tree extraction required around 20 s.The proposed method is able to accurately and efficiently determine the direction and radius of coronary arteries based on information derived directly from the image data. The method can be trained with limited training data, and once trained allows fast automatic or interactive extraction of coronary artery trees from CCTA images.     

Ultrasound imaging in three and four dimensions

Three-dimensional (3D) reconstruction of ultrasound images was first demonstrated nearly 15 years ago, but only now is becoming a clinical reality. In the meantime, methods for 3D reconstruction of CT and MRI images have achieved an advanced state of development, and 3D imaging with these modalities has been applied widely in clinical practice. 3D applications in ultrasound have lagged behind CT and MRI, because ultrasound data is much more difficult to render in 3D, for a variety of technical reasons, than either CT or MRI data. Only in the past few years has the computing power of ultrasound equipment reached a level adequate enough for the complex signal processing tasks needed to render ultrasound data in three dimensions. At this point in time, the clinical application of 3D ultrasound is likely to advance rapidly, as improved 3D rendering technology becomes more widely available. This article is a review of the present status of 3D ultrasound imaging. It begins by comparing the characteristics of CT, MRI, and ultrasound image data that either make these data amenable or not amenable to 3D reconstruction. The article then considers the technical features involved with acquiring an ultrasound 3D data set and the mechanisms for reconstructing the images. Finally, the article reviews the literature that is available regarding clinical application of 3D ultrasound in obstetrics, ultrasound, the abdomen, and blood vessels.     

Charisma: An integrated approach to automatic H&E-stained skeletal muscle cell segmentation using supervised learning and novel robust clump splitting

Histological image analysis plays a key role in understanding the effects of disease and treatment responses at the cellular level. However, evaluating histology images by hand is time-consuming and subjective. While semi-automatic and automatic approaches for image segmentation give acceptable results in some branches of histological image analysis, until now this has not been the case when applied to skeletal muscle histology images. We introduce Charisma, a new top-down cell segmentation framework for histology images which combines image processing techniques, a supervised trained classifier and a novel robust clump splitting algorithm. We evaluate our framework on real-world data from intensive care unit patients. Considering both segmentation and cell property distributions, the results obtained by our method correspond well to the ground truth, outperforming other examined methods.