The latest developments combining deep learning technology and medical image data have attracted wide attention and provide efficient noninvasive methods for the early diagnosis of breast cancer. The success of this task often depends on a large amount of data annotated by medical experts, which is time-consuming and may not always be feasible in the biomedical field. The lack of interpretability has greatly hindered the application of deep learning in the medical field. Currently, deep stable learning, including causal inference, make deep learning models more predictive and interpretable. In this study, to distinguish malignant tumors in Breast Imaging-Reporting and Data System (BI-RADS) category 3-4A breast lesions, we propose BD-StableNet, a deep stable learning model for the automatic detection of lesion areas. In this retrospective study, we collected 3103 breast ultrasound images (1418 benign and 1685 malignant lesions) from 493 patients (361 benign and 132 malignant lesion patients) for model training and testing. Compared with other mainstream deep learning models, BD-StableNet has better prediction performance (accuracy = 0.952, area under the curve (AUC) = 0.982, precision = 0.970, recall = 0.941, F1-score = 0.955 and specificity = 0.965). The lesion area prediction and class activation map (CAM) results both verify that our proposed model is highly interpretable. The results indicate that BD-StableNet significantly enhances diagnostic accuracy and interpretability, offering a promising noninvasive approach for the diagnosis of BI-RADS category 3-4A breast lesions. Clinically, the use of BD-StableNet could reduce unnecessary biopsies, improve diagnostic efficiency, and ultimately enhance patient outcomes by providing more precise and reliable assessments of breast lesions.

Cerebral arterial and venous flow (A/V) classification is a key parameter for understanding dynamic changes in neonatal brain perfusion. Currently, transfontanellar ultrasound Doppler imaging is the reference clinical technique able to discriminate between A/V using vascular indices such as resistivity index (RI) or pulsatility index (PI). However, under conditions of slow arterial and venular flow, small signal fluctuations can lead to potential misclassifications of vessels. Recently, ultrafast ultrasound imaging has paved the way for better sensitivity and spatial resolution. Here, we show that A/V classification can be performed robustly using ultrafast Doppler spectrogram. Approach: The overall classification steps are as follows: for any pixel within a vessel, a normalized Doppler spectrogram (NDS) is computed that allows for normalized correlation analysis with ground-truth signals that were established semi-automatically based on anatomical/physiological references. Furthermore, A/V classification is performed by computing Pearson correlation coefficient between NDS in ground-truth domains and the individual pixel's NDS inside vessels and finding an optimal threshold. Main Results: When applied to human newborns (n= 40), the overall accuracy, sensitivity, and specificity were found to be 88.5% ± 6.7%, 88.5% ± 6.5%, and 87.0% ± 8.8% respectively. We also examined strategies to fully automate this process, leading to a moderate decrease of 1%-3% in the same metrics. Additionally, when compared to the main clinical metrics such as RI, and PI, the receiver operating characteristic curves exhibited higher areas under the curve; on average by +36% (p < 0.0001) in the full imaging sector, +35% (p = 0.0116) in the cortical regions, +53% (p < 0.0001) in the basal ganglia, +28% (p = 0.0051) in the cingulate gyrus, and +35% (p < 0.0001) in the remaining brain structures. Significance: Our findings suggest that the proposed NDS-based approach can distinguish between A/V when studying cerebral perfusion in neonates. .

Channelized Hotelling model observers are efficient at simulating the human observer visual performance in medical imaging detection tasks. However, channelized Hotelling observers (CHO) are subject to statistical biases from zero-signal and finite-sample effects. The point estimate of the d' value is also not always symmetric with exact confidence interval (CI) bounds determined for the infinitely trained CHO. A method for correcting these statistical biases and CI asymmetry is studied.

To develop a nonlinear, model-based parameter estimation method directly from incomplete measurements in k-w space for robust spectral analysis in highly accelerated chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI).

. Low-dose computed tomography (LDCT) is an imaging technique that can effectively help patients reduce radiation dose, which has attracted increasing interest from researchers in the field of medical imaging. Nevertheless, LDCT imaging is often affected by a large amount of noise, making it difficult to clearly display subtle abnormalities or lesions. Therefore, this paper proposes a multiple complementary priors CT image reconstruction method by simultaneously considering both the internal prior and external image information of CT images, thereby enhancing the reconstruction quality of CT images.. Specifically, we propose a CT image reconstruction method based on weighted nonconvex low-rank regularized group sparse and deep image priors under hybrid plug-and-play framework by utilizing the weighted nonconvex low rankness and group sparsity of dictionary domain coefficients of each group of similar patches, and a convolutional neural network denoiser. To make the proposed reconstruction problem easier to tackle, we utilize the alternate direction method of multipliers for optimization.. To verify the performance of the proposed method, we conduct detailed simulation experiments on the images of the abdominal, pelvic, and thoracic at projection views of 45, 65, and 85, and at noise levels of1×105and1×106, respectively. A large number of qualitative and quantitative experimental results indicate that the proposed method has achieved better results in texture preservation and noise suppression compared to several existing iterative reconstruction methods.. The proposed method fully considers the internal nonlocal low rankness and sparsity, as well as the external local information of CT images, providing a more effective solution for CT image reconstruction. Consequently, this method enables doctors to diagnose and treat diseases more accurately by reconstructing high-quality CT images.

To develop a deep reinforcement learning (DRL) agent to self-interact with the treatment planning system (TPS) to automatically generate intensity modulated radiation therapy (IMRT) treatment plans for head-and-neck (HN) cancer with consistent organ-at-risk (OAR) sparing performance. Methods: With IRB approval, one hundred and twenty HN patients receiving IMRT were included. The DRL agent was trained with 20 patients. During each inverse optimization process, the intermediate dosimetric endpoints' value, dose volume constraints value and structure objective function loss were collected as the DRL states. By adjusting the objective constraints as actions, the agent learned to seek optimal rewards by balancing OAR sparing and planning target volume (PTV) coverage. Reward computed from current dose-volume-histogram (DVH) endpoints and clinical objectives were sent back to the agent to update action policy during model training. The trained agent was evaluated with the rest 100 patients. Results: The DRL agent was able to generate a clinically acceptable IMRT plan within 12.4±3.1 minutes without human intervention. DRL plans showed lower PTV maximum dose (109.2%) compared to clinical plans (112.4%) (p<.05). Average median dose of left parotid, right parotid, oral cavity, larynx, pharynx of DRL plans were 15.6Gy, 12.2Gy, 25.7Gy, 27.3Gy and 32.1Gy respectively, comparable to 17.1 Gy,15.7Gy, 24.4Gy, 23.7Gy and 35.5Gy of corresponding clinical plans. The maximum dose of cord+5mm, brainstem and mandible were also comparable between the two groups. In addition, DRL plans demonstrated reduced variability, as evidenced by smaller 95% confidence intervals. The total MU of the DRL plans was 1611 vs 1870 (p<.05) of clinical plans. The results signaled the DRL's consistent planning strategy compared to the planners' occasional back-and-forth decision-making during planning. Conclusion: The proposed deep reinforcement learning (DRL) agent is capable of efficiently generating HN IMRT plans with consistent quality. .

Positronium lifetime tomography (PLT) is an emerging modality that aims to reconstruct 3D images of positronium lifetime in humans and animals in vivo. The lifetime of ortho-positronium can be influenced by the microstructure and the concentration of bio-active molecules in tissue, providing valuable information for better understanding disease progression and treatment response. However, efficient high-resolution lifetime image reconstruction methods are currently lacking. Existing methods are either computationally intensive or have poor spatial resolution. This paper presents a fast, high-resolution lifetime image reconstruction method for positronium lifetime tomography. Approach: The proposed method, called SIMPLE-Moment (Statistical IMage reconstruction of Positron annihilation LifetimE by Moment weighting), first reconstructs a set of moment images and then estimates the ortho-positronium lifetime image using the method of moments. The implementation of SIMPLE-Moment requires minimal modification to the conventional ordered subset expectation maximization (OSEM) algorithm. Main results: With reasonable assumptions, the proposed method can reconstruct an ortho-positronium lifetime image with a computational cost equivalent to three standard PET image reconstructions. A Monte Carlo simulation study based on an existing time-of-flight (TOF) PET scanner demonstrates that the ortho-positronium lifetime image reconstructed by SIMPLE-Moment is accurate and comparable to results obtained using the more computationally intensive SPLIT method. Significance: The proposed SIMPLE-Moment method provides an efficient approach to high-resolution reconstruction of ortho-positronium lifetime images. By reducing computational costs while enhancing spatial resolution, this method has the potential to make positronium lifetime tomography more accessible and practical for clinical and research applications. .

Motivated by the diagnostic value of tissue viscosity beyond elasticity, the goal of this work is to develop robust methodologies based on shear wave elastography (SWE) to reconstruct combined elasticity and viscosity maps of soft tissues out of the measurement plane.

Conventionally, if two metabolic processes are of interest for image analysis, two separate, sequential PET scans are performed. However, sequential PET scans cannot simultaneously display the metabolic targets. The concurrent study of two simultaneous PET scans could provide new insights into the causes of diseases. Approach: In this work, we propose a reconstruction algorithm for the simultaneous injection of a β+-emitter emitting only annihilation photons and a β+-γ-emitter emitting annihilation photons and an additional prompt γ-photon. As in previous works, the γ-photon is used to identify events originating from the β+-γ-emitter. However, due to e.g. attenuation, the γ-photon is often not detected and not all events can correctly be associated with the β+-γ-emitter as they are detected as double coincidences. In contrast to previous works, we estimate this number of double coincidences with origin in the β+-γ, emitter including the attenuation of the prompt γ, and incorporate this estimation in the forward-projection of the ML-EM algorithm. For evaluation, we simulate different scenarios with varying objects and attenuation maps. The nuclide 18F serves as β+-emitter, while 44Sc functions as β+-γ emitter. The performance of the algorithm is assessed by calculating the residual error of the β+-γ-emitter in the reconstructed β+-emitter image. Additionally, the intensity values in the simulated cylinders of the ground truth (GT) and the reconstructed images are compared. Main Results: The remaining activity in the β+-emitter image varied from 0.4% to 3.7%. The absolute percentage difference between GT and reconstructed intensity for the pure β+ emitter images was found to be between 3.0 and 7.4% for all cases. The absolute percentage difference between GT and reconstructed intensity for the β+-γ emitter images ranged from 8.7 to 10.4% for all simulated cases. Significance: These results demonstrate that our approach can reconstruct two separate images with a good quantitation accurac.

We previously developed a FLASH planning framework for streamlined pin-ridge-filter (pin-RF) design, demonstrating its feasibility for single-energy proton FLASH planning. In this study, we refined the pin-RF design for easy assembly using reusable modules, focusing on its application in liver stereotactic ablative body radiotherapy (SABR). This framework generates an intermediate intensity-modulated proton therapy (IMPT) plan and translates it into step widths and thicknesses of pin-RFs for a single-energy FLASH plan. Parameters like energy spacing, monitor unit limit, and spot quantity were adjusted during IMPT planning, resulting in pin-RFs assembled using predefined modules with widths from 1 to 6 mm, each with a water-equivalent-thickness of 5 mm. This approach was validated on three liver SABR cases. FLASH doses, quantified using the FLASH effectiveness model at 1 to 5 Gy thresholds, were compared to conventional IMPT (IMPT-CONV) doses to assess clinical benefits. The highest demand for 6 mm width modules, moderate for 2-4 mm, and minimal for 1- and 5-mm modules were shown across all cases. At lower dose thresholds, the two-beam case reduced indicators including liver V21Gy and skin Dmax by >19.4%, while the three-beam cases showed reductions ≤11.4%, indicating the need for higher fractional beam doses for an enhanced FLASH effect. Positive clinical benefits were seen only in the two-beam case at the 5 Gy threshold. At the 1 Gy threshold, the two-beam FLASH plan outperformed the IMPT-CONV plan, reducing dose indicators for all relevant normal tissues by up to 31.2%. In contrast, the three-beam cases showed negative clinical benefits, with skin Dmax and liver V21Gy increasing by up to 17.4% due to lower fractional beam doses and closer beam arrangements. This study evaluated the feasibility of modularizing streamlined pin-RFs in single-energy proton FLASH planning for liver SABR, offering guidance on optimal module composition and strategies to enhance FLASH planning.

Contrast agents in CT scans can compromise the accuracy of dose calculations in radiation therapy planning, especially for particle therapy. This often requires an additional non-contrast CT scan, increasing radiation exposure and introducing potential registration errors. Our goal is to resolve these issues by accurately estimating radiotherapy parameters from dual virtual non-contrast (dual-VNC) images generated by contrast-enhanced dual-energy CT (DECT) scans, while accounting for noise and variability in tissue composition. Approach: A new Bayesian model is introduced to estimate dual-VNC Hounsfield units from contrast-enhanced DECT data. The model defines a prior distribution that describes tissue variations in terms of elemental compositions and mass densities. Multiple reference tissues are used to estimate variations across human tissues. A likelihood distribution is also defined to model the noise contained in CT data. The model is thoroughly validated in a simulated environment including 12 virtual patients under low and high iodine uptake scenarios, while incorporating noise and beam hardening effects. The eigentissue decomposition (ETD) technique is used to derive elemental compositions and parameters critical for radiotherapy from the dual-VNC images, such as electron density (ρ), particle stopping power (SPR), and photon energy absorption coefficient (EAC) Main results: The proposed method yields accurate voxelwise estimations for ρ, SPR, and EAC, with root mean square errors of 3.09%, 3.14%, and 1.34% for highly-enhanced tissues, compared to 5.93%, 6.39%, and 17.11% when the presence of contrast agent is ignored. It also demonstrates robustness to systematic shifts in tissue composition and bandwidth variations in the prior distribution, resulting in overall uncertainties down to 1.13%, 1.33%, and 0.86% for ρ, SPR, and EAC in soft tissues; 1.17%, 1.32%, and 1.34% in enhanced soft tissues; and 4.34%, 4.00%, and 2.50% in bone. Significance: The proposed method accurately derives radiotherapy parameters from contrast-enhanced DECT data and demonstrates robustness against systematic errors in reference data, highlighting its potential for clinical use.

The need for increased sensitivity in magnetic resonance imaging (MRI) is crucial for its advancement as an imaging modality. The development of passive Lenz Resonators for effective RF magnetic field focusing will improve MRI sensitivity via local amplification of MRI signal, thereby leading to more efficient diagnosis and patient treatment.

Objective - Microdosimetry is gaining increasing interest in particle therapy. Thanks to the advancements in microdosimeter technologies and the increasing number of experimental studies carried out in hadron therapy frameworks, it is proving to be a reliable experimental technique for radiation quality characterisation, quality assurance, and radiobiology studies. However, considering the variety of detectors used for microdosimetry, it is important to ensure the consistency of microdosimetric results measured with different types of microdosimeters. Approach - This work presents a novel multi-thickness microdosimeter and a methodology to characterise the radiation quality of a clinical carbon-ion beam. The novel device is a diamond detector made of three sensitive volumes (SV) of different thicknesses: 2, 6 and 12 μm. The SVs, which operate simultaneously, were accurately aligned and laterally positioned within 3mm. This allignement allowed for a comparison of the results with a negligible impact of the SVs alignment and their lateral positioning, ensuring the homogeneity of the measured radiation quality. An experimental campaign was carried out at MedAustron using a carbon-ion beam of typical clinical energy (284.7MeV/u). Main results - The measurement results allowed for a meticulous interpretation of its radiation quality, highlighting the effect of the SV thickness. The consistency of the microdosimetric spectra measured by detectors of different thicknesses is discussed by critically analysing the spectra and the differences observed. Significance - The methodology presented will be highly valuable for future experiments investigating the effects of the target volume size in radiobiology and could be easily adapted to the other particles employed in hadron therapy for clinical (i.e. protons) and for research purposes (e.g. helium, lithium and oxygen ions).

The arrangement of chromosomes in the cell nucleus has implications for cell radiosensitivity. The development of new tools to utilize Hi-C chromosome conformation data in nanoscale radiation track structure simulations allows for in silico investigation of this phenomenon. We have developed a framework employing Hi-C-based cell nucleus models in Monte Carlo radiation simulations, in conjunction with mechanistic models of DNA repair, to predict not only the initial radiation-induced DNA damage, but also the repair outcomes resulting from this damage, allowing us to investigate the role chromosome conformation plays in the biological outcome of radiation exposure. Approach: In this study, we used this framework to generate cell nucleus models based on Hi-C data from fibroblast and lymphoblastoid cells and explore the effects of cell type-specific chromosome structure on radiation response. The models were used to simulate external beam irradiation including DNA damage and subsequent DNA repair. The kinetics of the simulated DNA repair were compared with previous results. Main Results: We found that the fibroblast models resulted in a higher rate of inter-chromosome misrepair than the lymphoblastoid model, despite having similar amounts of initial DNA damage and total misrepairs for each irradiation scenario. Significance: This framework represents a step forward in radiobiological modeling and simulation allowing for more realistic investigation of radiosensitivity in different types of cells.

Plastic scintillation dosimeters (PSDs) are highly suitable for real-time dosimetry on the MR-linac. For optimal performance, the primary signal (scintillation) needs to be separated from secondary optical effects (Cerenkov, fluorescence and optical fiber attenuation). This requires a spectral separation approach and careful calibration. Currently, the 'classic' calibration is a multi-step procedure using both kV and MV X-ray sources, requiring an uninterrupted optical connection between the dosimeter and read-out system, complicating efficient use of PSDs. Therefore, we present a more time-efficient and more practical novel calibration technique for PSDs suitable for MR-linac dosimetry. Approach: The novel calibration relies on prior spectral information combined with two 10x10 cmfield irradiations on the 1.5 T MR-linac. Performance of the novel calibration technique was evaluated focusing on its reproducibility, performance characteristics (repeatability, linearity, dose rate dependency, output factors, angular response and detector angle dependency) and IMRT deliveries. To investigate the calibration stability over time, prior spectral information up to 315 days old was used. To quantify the time efficiency, each step of the novel and classic calibration was timed. Main results: The novel calibration showed a high reproducibility with a maximum relative standard deviation of 0.3%. The novel method showed maximum differences of 1.2% compared to the gold-standard calibration, while reusing old classic calibrations after reconnecting fibers showed differences up to 3.0%. The novel calibration improved time efficiency from 105 to 30 minutes compared to the classic method. Significance: The novel calibration method showed a gain in time efficiency and practicality while preserving the dosimetric accuracy. Therefore, this method can replace the traditional method for PSDs suitable for MR-linac dosimetry, using prior spectral information of up to a year. This novel calibration facilitates reconnecting the detector to the read-out system which would lead to unacceptable dosimetric results with the classic calibration method.

Magnetic resonance imaging (MRI) provides detailed structural information of the internal body organs and soft tissue regions of a patient in clinical diagnosis for disease detection, localization, and progress monitoring. MRI scanner hardware manufacturers incorporate various post-acquisition image-processing techniques into the scanner's computer software tools for different post-processing tasks. These tools provide a final image of adequate quality and essential features for accurate clinical reporting and predictive interpretation for better treatment planning. Different post-acquisition image-processing tasks for MRI quality enhancement include noise removal, motion artifact reduction, magnetic bias field correction, and eddy electric current effect removal. Recently, deep learning (DL) methods have shown great success in many research fields, including image and video applications. DL-based data-driven feature-learning approaches have great potential for MR image denoising and image-quality-degrading artifact correction. Recent studies have demonstrated significant improvements in image-analysis tasks using DL-based convolutional neural network (CNN) techniques. The promising capabilities and performance of DL techniques in various problem-solving domains have motivated researchers to adapt DL methods to medical image analysis and quality enhancement tasks. This paper presents a comprehensive review of DL-based state-of-the-art MRI quality enhancement and artifact removal methods for regenerating high-quality images while preserving essential anatomical and physiological feature maps without destroying important image information. Existing research gaps and future directions have also been provided by highlighting potential research areas for future developments, along with their importance and advantages in medical imaging.

The purpose of this study is to analytically derive and validate a novel radiation energy conservation principle for dose mapping via DIR. Approach: A radiation energy conservation principle for the DIR-based dose-deforming process was theoretically derived with a consideration of the volumetric Jacobian and proven using synthetic examples and a patient case. Furthermore, an energy difference error was proposed that can be used to evaluate the DIR-based dose accumulation uncertainty. For the analytical validation of the proposed energy conservation principle, a synthetic isotropic deformation was considered, and artificial deformation uncertainties were introduced. For the validation with a patient case, a ground truth set of CT images and the corresponding deformation was generated. Radiation energy calculation was performed using both the ground truth deformation and another deformation with uncertainty. Main results: The suggested energy conservation principle was preserved with uncertainty-free deformation, but not with error-containing deformations using both the synthetic examples and the patient case. For a synthetic example with a tumor volume reduction of 27.1% (10% reduction in length in all directions), the energy difference error was calculated to be -29.8% and 37.2% for an over-deforming and under-deforming DIR uncertainty of 0.3 cm. The energy difference error for the patient case (tumor volume reduction of 37.6%) was 2.9% for a displacement vector field with a registration error of 2.0 ± 3.2 mm. Significance: A novel energy conservation principle for DIR-based dose deformation and the corresponding energy difference error were mathematically formulated and successfully validated using simple synthetic examples and a patient example. With a consideration of the volumetric Jacobian, this investigation proposed a radiation energy conservation principle which can be met only with uncertainty-free deformations.

Liver cancer patients may benefit from proton therapy through increase of the tumor control probability (TCP). However, proton therapy is a limited resource and may not be available for all patients. We consider combined proton-photon liver SBRT treatments (CPPT) where only some fractions are delivered with protons. It is investigated how limited proton fractions can be used best for individual patients and optimally allocated within a patient group. Approach: Photon and proton treatment plans were created for five liver cancer patients. In CPPT, limited proton fractions may be optimally exploited by increasing the fraction dose compared to photon fraction dose. To determine a patient's optimal proton and photon fraction dose, we maximize the target BED while constraining the mean normal liver BED, which leads to an up- or downscaling of the proton and photon plan, respectively. The resulting CPPT balances the benefits of fractionation in the normal liver versus exploiting the superior proton dose distributions. After converting the target BED to TCP, the optimal number of proton fractions per patient is determined by maximizing the overall TCP of the patient group. Main results: For the individual patient, a CPPT treatment that delivers a higher fraction dose with protons than photons allows for dose escalation in the target compared to delivering the same proton and photon fraction dose. On the level of a patient group, CPPT may allow to distribute limited proton slots over several patients. Through an optimal use and allocation of proton fractions, CPPT may increase the average patient group TCP compared to a proton patient selection strategy where patients receive single-modality proton or photon treatments. Significance: Limited proton resources can be optimally exploited via CPPT by increasing the target dose in proton fractions and allocating available proton slots to patients with the highest TCP increase. .

In endodontic therapy, 3D Cone Beam Computerized Tomography (CBCT) and oral scan fusion models allow exact root canal channels and guidance. However, the point cloud model from CBCT has few data points and poor model features, limiting 3D fusion with oral scan data. Our aim to build a sub-regional point cloud resampling method and evaluate the precision of merging it with three-dimensional oral scan data. Approach: Two molars and four incisors were resampled for this investigation. Based on point cloud density and curvature, the rebuilt model was separated into the crown and cervical cavities. Using crown surface morphology, Divergence Index (DI) was employed to determine resampling points based on point dispersion. Improved Euclidean Clustering Rule (IECR) downsamples each point using its weight and joins the two halves using Iterative Nearest Neighbour (ICP) to create a complete resampled point cloud. After aligning with the oral scanning model, the maximum error, maximum distance, average distance, and other characteristics are calculated to assess resampling. Additionally, a cross-entropy kernel-based point cloud reconstruction depth selection method is given to determine the appropriate reconstruction depth. Main results: Applying the DI-IECR technique reduces the average distance between the resampled tooth point cloud and the point cloud generated by the dental scanner by around 20%. The maximum error remains same to that of the widely used method. This study also demonstrates that the use of the DI-IECR approach guarantees the complete representation of the coronal characteristics of the resampled reconstructed 3D model, rather than excessively focusing processing resources on pertinent but insignificant areas. Significance: Point cloud data and crown features are balanced using DI-IECR. When registered with the oral scan model, CBCT-generated point clouds are more accurate and timely, making them a better intraoperative navigation model.

Mild cognitive impairment (MCI) is a precursor stage of dementia characterized by mild cognitive decline in one or more cognitive domains, without meeting the criteria for dementia. MCI is considered a prodromal form of Alzheimer's disease (AD). Early identification of MCI is crucial for both intervention and prevention of AD. To accurately identify MCI, a novel multimodal 3D imaging data integration graph convolutional network (GCN) model is designed in this paper.The proposed model utilizes 3D-VGGNet to extract three-dimensional features from multimodal imaging data (such as structural magnetic resonance imaging and fluorodeoxyglucose positron emission tomography), which are then fused into feature vectors as the node features of a population graph. Non-imaging features of participants are combined with the multimodal imaging data to construct a population sparse graph. Additionally, in order to optimize the connectivity of the graph, we employed the pairwise attribute estimation （PAE） method to compute the edge weights based on non-imaging data, thereby enhancing the effectiveness of the graph structure. Subsequently, a population-based GCN integrates the structural and functional features of different modal images into the features of each participant for MCI classification.Experiments on the AD Neuroimaging Initiative demonstrated accuracies of 98.57%, 96.03%, and 96.83% for the normal controls (NC)-early MCI (EMCI), NC-late MCI (LMCI), and EMCI-LMCI classification tasks, respectively. The AUC, specificity, sensitivity, and F1-score are also superior to state-of-the-art models, demonstrating the effectiveness of the proposed model. Furthermore, the proposed model is applied to the ABIDE dataset for autism diagnosis, achieving an accuracy of 91.43% and outperforming the state-of-the-art models, indicating excellent generalization capabilities of the proposed model.This study demonstratethe proposed model's ability to integrate multimodal imaging data and its excellent ability to recognize MCI. This will help achieve early warning for AD and intelligent diagnosis of other brain neurodegenerative diseases.

Diffusion equation imaging processing is promising to enhance images showing lesions of bone metastasis (LBM). The Perona-Malik diffusion (PMD) model, which has been widely used and studied, is an anisotropic diffusion processing method to denoise or extract objects from an image effectively. However, the smoothing characteristics of PMD or its related method hinder extraction and enhancement of soft tissue regions of medical image such as computed tomography (CT), typically leaving an indistinct region with ambient tissues. Moreover, PMD expands the border region of the objects. A novel diffusion methodology must be used to enhance the LBM region effectively. Approach. For this study, we originally developed a diffusion equation quantification (DEQ) method that uses a filter function to selectively provide modulated diffusion according to the original locations of objects in an image. The structural similarity index measure (SSIM) and Lie derivative image analysis (LDIA) L-value map were used to evaluate image quality and processing. Main results. We determined superellipse function with its order n=4 for the LBM region. DEQ was found to be more effective at contrasting LBM for various LBM CT images than PMD or its improved models. DEQ yields enhancement agreeing with the indications of positron emission tomography despite complex lesions of bone metastasis comprising osteoblastic, osteoclastic, mixed tissues, and metal artifacts, which is innovative. Moreover, DEQ retained high quality of image (SSIM > 0.95), and achieved a low mean value of the L-value (< 0.001), indicative of our intended selective diffusion compared to other PMD models. Significance. Our method improved the visibility of mixed tissue lesions, which can assist computer visional framework and can help radiologists to produce accurate diagnose of LBM regions which are frequently overlooked in radiology findings because of the various degrees of visibility in CT images.