INTRODUCTION: The histopathological classification for antineutrophil cytoplasmic autoantibody (ANCA)-associated glomerulonephritis (ANCA-GN) is a well-established tool to reflect the variety of patterns and severity of lesions that can occur in kidney biopsies. It was demonstrated previously that deep learning (DL) approaches can aid in identifying histopathological classes of kidney diseases; for example, of diabetic kidney disease. These models can potentially be used as decision support tools for kidney pathologists. Although they reach high prediction accuracies, their "black box" structure makes them nontransparent. Explainable (X) artificial intelligence (AI) techniques can be used to make the AI model decisions accessible for human experts. We have developed a DL-based model, which detects and classifies the glomerular lesions according to the Berden classification. METHODS: Kidney biopsy slides of 80 patients with ANCA-GN from 3 European centers, who underwent a diagnostic kidney biopsy between 1991 and 2011, were included. We also investigated the explainability of our model using Gradient-weighted Class Activation Mapping (Grad-CAM) heatmaps. These maps were analyzed by pathologists to compare the decision-making criteria of humans and the DL model and assess the impact of different training settings. RESULTS: The DL model shows a prediction accuracy of 93% for classifying lesions. The heatmaps from our trained DL models showed that the most predictive areas in the image correlated well with the areas deemed to be important by the pathologist. CONCLUSION: We present the first DL-based computational pipeline for classifying ANCA-GN kidney biopsies as per the Berden classification. XAI techniques helped us to make the decision-making criteria of the DL accessible for renal pathologists, potentially improving clinical decision-making.
- Publication type
- Journal Article MeSH
BACKGROUND: Neuromuscular diseases (NMDs) are rare disorders characterized by progressive muscle fibre loss, leading to replacement by fibrotic and fatty tissue, muscle weakness and disability. Early diagnosis is critical for therapeutic decisions, care planning and genetic counselling. Muscle magnetic resonance imaging (MRI) has emerged as a valuable diagnostic tool by identifying characteristic patterns of muscle involvement. However, the increasing complexity of these patterns complicates their interpretation, limiting their clinical utility. Additionally, multi-study data aggregation introduces heterogeneity challenges. This study presents a novel multi-study harmonization pipeline for muscle MRI and an AI-driven diagnostic tool to assist clinicians in identifying disease-specific muscle involvement patterns. METHODS: We developed a preprocessing pipeline to standardize MRI fat content across datasets, minimizing source bias. An ensemble of XGBoost models was trained to classify patients based on intramuscular fat replacement, age at MRI and sex. The SHapley Additive exPlanations (SHAP) framework was adapted to analyse model predictions and identify disease-specific muscle involvement patterns. To address class imbalance, training and evaluation were conducted using class-balanced metrics. The model's performance was compared against four expert clinicians using 14 previously unseen MRI scans. RESULTS: Using our harmonization approach, we curated a dataset of 2961 MRI samples from genetically confirmed cases of 20 paediatric and adult NMDs. The model achieved a balanced accuracy of 64.8% ± 3.4%, with a weighted top-3 accuracy of 84.7% ± 1.8% and top-5 accuracy of 90.2% ± 2.4%. It also identified key features relevant for differential diagnosis, aiding clinical decision-making. Compared to four expert clinicians, the model obtained the highest top-3 accuracy (75.0% ± 4.8%). The diagnostic tool has been implemented as a free web platform, providing global access to the medical community. CONCLUSIONS: The application of AI in muscle MRI for NMD diagnosis remains underexplored due to data scarcity. This study introduces a framework for dataset harmonization, enabling advanced computational techniques. Our findings demonstrate the potential of AI-based approaches to enhance differential diagnosis by identifying disease-specific muscle involvement patterns. The developed tool surpasses expert performance in diagnostic ranking and is accessible to clinicians worldwide via the Myo-Guide online platform.
- MeSH
- Adult MeSH
- Internet MeSH
- Middle Aged MeSH
- Humans MeSH
- Magnetic Resonance Imaging * methods MeSH
- Neuromuscular Diseases * diagnosis diagnostic imaging MeSH
- Machine Learning * MeSH
- Check Tag
- Adult MeSH
- Middle Aged MeSH
- Humans MeSH
- Male MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
Achieving a reliable and accurate biomedical image segmentation is a long-standing problem. In order to train or adapt the segmentation methods or measure their performance, reference segmentation masks are required. Usually gold-standard annotations, i.e. human-origin reference annotations, are used as reference although they are very hard to obtain. The increasing size of the acquired image data, large dimensionality such as 3D or 3D + time, limited human expert time, and annotator variability, typically result in sparsely annotated gold-standard datasets. Reliable silver-standard annotations, i.e. computer-origin reference annotations, are needed to provide dense segmentation annotations by fusing multiple computer-origin segmentation results. The produced dense silver-standard annotations can then be either used as reference annotations directly, or converted into gold-standard ones with much lighter manual curation, which saves experts' time significantly. We propose a novel full-resolution multi-rater fusion convolutional neural network (CNN) architecture for biomedical image segmentation masks, called DeepFuse, which lacks any down-sampling layers. Staying everywhere at the full resolution enables DeepFuse to fully benefit from the enormous feature extraction capabilities of CNNs. DeepFuse outperforms the popular and commonly used fusion methods, STAPLE, SIMPLE and other majority-voting-based approaches with statistical significance on a wide range of benchmark datasets as demonstrated on examples of a challenging task of 2D and 3D cell and cell nuclei instance segmentation for a wide range of microscopy modalities, magnifications, cell shapes and densities. A remarkable feature of the proposed method is that it can apply specialized post-processing to the segmentation masks of each rater separately and recover under-segmented object parts during the refinement phase even if the majority of inputs vote otherwise. Thus, DeepFuse takes a big step towards obtaining fast and reliable computer-origin segmentation annotations for biomedical images.
- MeSH
- Humans MeSH
- Neural Networks, Computer * MeSH
- Image Processing, Computer-Assisted * methods MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
BACKGROUND: This study aims to evaluate the feasibility of generating pseudo-normal single photon emission computed tomography (SPECT) data from potentially abnormal images. These pseudo-normal images are primarily intended for use in an on-the-fly data harmonization technique. MATERIAL AND METHODS: The methodology was tested on brain SPECT with [123I]Ioflupane. The proposed model for generating a pseudo-normal image was based on a variational autoencoder (VAE) designed to process 2D sinograms of the brain [123I]-FP-CIT SPECT, potentially exhibiting abnormal uptake. The model aimed to predict SPECT sinograms with corresponding normal uptake. Training, validation, and testing datasets were created by SPECT simulator from 45 brain masks segmented from real patient's magnetic resonance (MR) scans, using various uptake levels. The training and validation datasets each comprised 612 and 360 samples, respectively, drawn from 36 brain masks. The testing dataset contained 153 samples based on 9 brain masks. VAE performance was evaluated through brain dimensions, Dice similarity coefficient (DSC) and specific binding ratio. RESULTS: Mean DSC was 80% for left basal ganglia and 84% for right basal ganglia. The proposed VAE demonstrated excellent consistency in predicting basal ganglia shape, with a coefficient of variation of DSC being less than 1.1%. CONCLUSIONS: The study demonstrates that VAE can effectively estimate an individualized pseudo-normal distribution of the radiotracer [123I]-FP-CIT SPECT from abnormal SPECT images. The main limitations of this preliminary research are the limited availability of real brain MR data, used as input for the SPECT data simulator, and the simplified simulation setup employed to create the synthetic dataset.
- MeSH
- Tomography, Emission-Computed, Single-Photon * MeSH
- Humans MeSH
- Brain * diagnostic imaging MeSH
- Image Processing, Computer-Assisted * MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
BACKGROUND: This study develops a deep learning-based automated lesion segmentation model for whole-body 3D18F-fluorodeoxyglucose (FDG)-Position emission tomography (PET) with computed tomography (CT) images agnostic to disease location and site. METHOD: A publicly available lesion-annotated dataset of 1014 whole-body FDG-PET/CT images was used to train, validate, and test (70:10:20) eight configurations with 3D U-Net as the backbone architecture. The best-performing model on the test set was further evaluated on 3 different unseen cohorts consisting of osteosarcoma or neuroblastoma (OS cohort) (n = 13), pediatric solid tumors (ST cohort) (n = 14), and adult Pheochromocytoma/Paraganglioma (PHEO cohort) (n = 40). Both lesion-level and patient-level statistical analyses were conducted to validate the performance of the model on different cohorts. RESULTS: The best performing 3D full resolution nnUNet model achieved a lesion-level sensitivity and DISC of 71.70 % and 0.40 for the test set, 97.83 % and 0.73 for ST, 40.15 % and 0.36 for OS, and 78.37 % and 0.50 for the PHEO cohort. For the test set and PHEO cohort, the model has missed small volume and lower uptake lesions (p < 0.01), whereas no statistically significant differences (p > 0.05) were found in the false positive (FP) and false negative lesions volume and uptake for the OS and ST cohort. The predicted total lesion glycolysis is slightly higher than the ground truth because of FP calls, which experts can easily check and reject. CONCLUSION: The developed deep learning-based automated lesion segmentation AI model which utilizes 3D_FullRes configuration of the nnUNet framework showed promising and reliable performance for the whole-body FDG-PET/CT images.
- MeSH
- Whole Body Imaging * methods MeSH
- Deep Learning * MeSH
- Child MeSH
- Adult MeSH
- Fluorodeoxyglucose F18 * MeSH
- Cohort Studies MeSH
- Middle Aged MeSH
- Humans MeSH
- Adolescent MeSH
- Neoplasms * diagnostic imaging MeSH
- Positron Emission Tomography Computed Tomography * methods MeSH
- Image Processing, Computer-Assisted * methods MeSH
- Check Tag
- Child MeSH
- Adult MeSH
- Middle Aged MeSH
- Humans MeSH
- Adolescent MeSH
- Male MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
- Validation Study MeSH
At a population level, the European Society of Gastrointestinal Endoscopy (ESGE), the European Helicobacter and Microbiota Study Group (EHMSG), and the European Society of Pathology (ESP) suggest endoscopic screening for gastric cancer (and precancerous conditions) in high-risk regions (age-standardized rate [ASR] > 20 per 100 000 person-years) every 2 to 3 years or, if cost-effectiveness has been proven, in intermediate risk regions (ASR 10-20 per 100 000 person-years) every 5 years, but not in low-risk regions (ASR < 10).ESGE/EHMSG/ESP recommend that irrespective of country of origin, individual gastric risk assessment and stratification of precancerous conditions is recommended for first-time gastroscopy. ESGE/EHMSG/ESP suggest that gastric cancer screening or surveillance in asymptomatic individuals over 80 should be discontinued or not started, and that patients' comorbidities should be considered when treatment of superficial lesions is planned.ESGE/EHMSG/ESP recommend that a high quality endoscopy including the use of virtual chromoendoscopy (VCE), after proper training, is performed for screening, diagnosis, and staging of precancerous conditions (atrophy and intestinal metaplasia) and lesions (dysplasia or cancer), as well as after endoscopic therapy. VCE should be used to guide the sampling site for biopsies in the case of suspected neoplastic lesions as well as to guide biopsies for diagnosis and staging of gastric precancerous conditions, with random biopsies to be taken in the absence of endoscopically suspected changes. When there is a suspected early gastric neoplastic lesion, it should be properly described (location, size, Paris classification, vascular and mucosal pattern), photodocumented, and two targeted biopsies taken.ESGE/EHMSG/ESP do not recommend routine performance of endoscopic ultrasonography (EUS), computed tomography (CT), magnetic resonance imaging (MRI), or positron emission tomography (PET)-CT prior to endoscopic resection unless there are signs of deep submucosal invasion or if the lesion is not considered suitable for endoscopic resection.ESGE/EHMSG/ESP recommend endoscopic submucosal dissection (ESD) for differentiated gastric lesions clinically staged as dysplastic (low grade and high grade) or as intramucosal carcinoma (of any size if not ulcerated or ≤ 30 mm if ulcerated), with EMR being an alternative for Paris 0-IIa lesions of size ≤ 10 mm with low likelihood of malignancy.ESGE/EHMSG/ESP suggest that a decision about ESD can be considered for malignant lesions clinically staged as having minimal submucosal invasion if differentiated and ≤ 30 mm; or for malignant lesions clinically staged as intramucosal, undifferentiated and ≤ 20 mm; and in both cases with no ulcerative findings.ESGE/EHMSG/ESP recommends patient management based on the following histological risk after endoscopic resection: Curative/very low-risk resection (lymph node metastasis [LNM] risk < 0.5 %-1 %): en bloc R0 resection; dysplastic/pT1a, differentiated lesion, no lymphovascular invasion, independent of size if no ulceration and ≤ 30 mm if ulcerated. No further staging procedure or treatment is recommended.Curative/low-risk resection (LNM risk < 3 %): en bloc R0 resection; lesion with no lymphovascular invasion and: a) pT1b, invasion ≤ 500 μm, differentiated, size ≤ 30 mm; or b) pT1a, undifferentiated, size ≤ 20 mm and no ulceration. Staging should be completed, and further treatment is generally not necessary, but a multidisciplinary discussion is required. Local-risk resection (very low risk of LNM but increased risk of local persistence/recurrence): Piecemeal resection or tumor-positive horizontal margin of a lesion otherwise meeting curative/very low-risk criteria (or meeting low-risk criteria provided that there is no submucosal invasive tumor at the resection margin in the case of piecemeal resection or tumor-positive horizontal margin for pT1b lesions [invasion ≤ 500 μm; well-differentiated; size ≤ 30 mm, and VM0]). Endoscopic surveillance/re-treatment is recommended rather than other additional treatment. High-risk resection (noncurative): Any lesion with any of the following: (a) a positive vertical margin (if carcinoma) or lymphovascular invasion or deep submucosal invasion (> 500 μm from the muscularis mucosae); (b) poorly differentiated lesions if ulceration or size > 20 mm; (c) pT1b differentiated lesions with submucosal invasion ≤ 500 μm with size > 30 mm; or (d) intramucosal ulcerative lesion with size > 30 mm. Complete staging and strong consideration for additional treatments (surgery) in multidisciplinary discussion.ESGE/EHMSG/ESP suggest the use of validated endoscopic classifications of atrophy (e. g. Kimura-Takemoto) or intestinal metaplasia (e. g. endoscopic grading of gastric intestinal metaplasia [EGGIM]) to endoscopically stage precancerous conditions and stratify the risk for gastric cancer.ESGE/EHMSG/ESP recommend that biopsies should be taken from at least two topographic sites (2 biopsies from the antrum/incisura and 2 from the corpus, guided by VCE) in two separate, clearly labeled vials. Additional biopsy from the incisura is optional.ESGE/EHMSG/ESP recommend that patients with extensive endoscopic changes (Kimura C3 + or EGGIM 5 +) or advanced histological stages of atrophic gastritis (severe atrophic changes or intestinal metaplasia, or changes in both antrum and corpus, operative link on gastritis assessment/operative link on gastric intestinal metaplasia [OLGA/OLGIM] III/IV) should be followed up with high quality endoscopy every 3 years, irrespective of the individual's country of origin.ESGE/EHMSG/ESP recommend that no surveillance is proposed for patients with mild to moderate atrophy or intestinal metaplasia restricted to the antrum, in the absence of endoscopic signs of extensive lesions or other risk factors (family history, incomplete intestinal metaplasia, persistent H. pylori infection). This group constitutes most individuals found in clinical practice.ESGE/EHMSG/ESP recommend H. pylori eradication for patients with precancerous conditions and after endoscopic or surgical therapy.ESGE/EHMSG/ESP recommend that patients should be advised to stop smoking and low-dose daily aspirin use may be considered for the prevention of gastric cancer in selected individuals with high risk for cardiovascular events.
- MeSH
- Biopsy MeSH
- Early Detection of Cancer * methods standards MeSH
- Gastroscopy * standards MeSH
- Risk Assessment MeSH
- Helicobacter Infections complications MeSH
- Humans MeSH
- Stomach Neoplasms * pathology diagnosis therapy MeSH
- Precancerous Conditions * pathology diagnosis therapy MeSH
- Societies, Medical MeSH
- Gastric Mucosa pathology diagnostic imaging MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
- Practice Guideline MeSH
- Geographicals
- Europe MeSH
BACKGROUND: Evidence suggests that brain-computer interface (BCI)-based rehabilitation strategies show promise in overcoming the limited recovery potential in the chronic phase of stroke. However, the specific mechanisms driving motor function improvements are not fully understood. OBJECTIVE: We aimed at elucidating the potential functional brain connectivity changes induced by BCI training in participants with chronic stroke. METHODS: A longitudinal crossover design was employed with two groups of participants over the span of 4 weeks to allow for within-subject (n = 21) and cross-group comparisons. Group 1 (n = 11) underwent a 6-day motor imagery-based BCI training during the second week, whereas Group 2 (n = 10) received the same training during the third week. Before and after each week, both groups underwent resting state functional MRI scans (4 for Group 1 and 5 for Group 2) to establish a baseline and monitor the effects of BCI training. RESULTS: Following BCI training, an increased functional connectivity was observed between the medial prefrontal cortex of the default mode network (DMN) and motor-related areas, including the premotor cortex, superior parietal cortex, SMA, and precuneus. Moreover, these changes were correlated with the increased motor function as confirmed with upper-extremity Fugl-Meyer assessment scores, measured before and after the training. CONCLUSIONS: Our findings suggest that BCI training can enhance brain connectivity, underlying the observed improvements in motor function. They provide a basis for developing novel rehabilitation approaches using non-invasive brain stimulation for targeting functionally relevant brain regions, thereby augmenting BCI-induced neuroplasticity and enhancing motor recovery.
- MeSH
- Stroke * physiopathology diagnostic imaging MeSH
- Chronic Disease MeSH
- Default Mode Network * physiopathology diagnostic imaging MeSH
- Adult MeSH
- Cross-Over Studies MeSH
- Middle Aged MeSH
- Humans MeSH
- Longitudinal Studies MeSH
- Magnetic Resonance Imaging MeSH
- Brain * physiopathology diagnostic imaging MeSH
- Nerve Net * physiopathology diagnostic imaging MeSH
- Stroke Rehabilitation * methods MeSH
- Brain-Computer Interfaces * MeSH
- Aged MeSH
- Check Tag
- Adult MeSH
- Middle Aged MeSH
- Humans MeSH
- Male MeSH
- Aged MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
BACKGROUND: Advancements in artificial intelligence (AI) and machine learning (ML) have revolutionized the medical field and transformed translational medicine. These technologies enable more accurate disease trajectory models while enhancing patient-centered care. However, challenges such as heterogeneous datasets, class imbalance, and scalability remain barriers to achieving optimal predictive performance. METHODS: This study proposes a novel AI-based framework that integrates Gradient Boosting Machines (GBM) and Deep Neural Networks (DNN) to address these challenges. The framework was evaluated using two distinct datasets: MIMIC-IV, a critical care database containing clinical data of critically ill patients, and the UK Biobank, which comprises genetic, clinical, and lifestyle data from 500,000 participants. Key performance metrics, including Accuracy, Precision, Recall, F1-Score, and AUROC, were used to assess the framework against traditional and advanced ML models. RESULTS: The proposed framework demonstrated superior performance compared to classical models such as Logistic Regression, Random Forest, Support Vector Machines (SVM), and Neural Networks. For example, on the UK Biobank dataset, the model achieved an AUROC of 0.96, significantly outperforming Neural Networks (0.92). The framework was also efficient, requiring only 32.4 s for training on MIMIC-IV, with low prediction latency, making it suitable for real-time applications. CONCLUSIONS: The proposed AI-based framework effectively addresses critical challenges in translational medicine, offering superior predictive accuracy and efficiency. Its robust performance across diverse datasets highlights its potential for integration into real-time clinical decision support systems, facilitating personalized medicine and improving patient outcomes. Future research will focus on enhancing scalability and interpretability for broader clinical applications.
- MeSH
- Databases, Factual MeSH
- Humans MeSH
- Neural Networks, Computer MeSH
- Patient-Centered Care * MeSH
- Machine Learning * MeSH
- Translational Science, Biomedical MeSH
- Translational Research, Biomedical MeSH
- Artificial Intelligence * MeSH
- Treatment Outcome MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
OBJECTIVES: This evidence implementation project aimed to assess and improve compliance with evidence-based neuroimaging criteria for adult patients with suspected stroke. INTRODUCTION: Stroke is the second leading cause of mortality and severe disability, requiring timely and accurate diagnosis. Clinical guidelines recommend brain imaging within 60 minutes of hospital arrival for suspected stroke patients. This project involved hospitals in North West Anglia NHS Foundation Trust, UK, serving 850,000 people with over 800 admissions annually. METHODS: The JBI Evidence Implementation Framework was used to guide this project. JBI software, the Practical Application of Clinical Evidence System (PACES), as well as JBI's Getting Research into Practice (GRiP) approach, were used to conduct the audit and implementation phases. The project followed three stages: (1) implementation planning, (2) baseline assessment and implementation, and (3) impact evaluation. Three audit criteria were used to represent best practices for diagnosing suspected stroke patients. RESULTS: The baseline audit revealed low compliance with the first criterion, with only 2.9% (1/35) of patients receiving a CT head scan within 1 hour of admission. In the follow-up audit, compliance improved to 45.2% (14/31). The other two criteria, diagnosis by a trained health care professional and baseline ECG assessment, had already achieved 100% compliance in the baseline audit. CONCLUSIONS: Compliance with evidence-based neuroimaging criteria improved after implementing targeted educational strategies and training. The rate of CT scans conducted within 1 hour increased, although door-to-imaging times remain suboptimal compared with achievable benchmarks of ≤ 20 minutes. Ongoing education and training are crucial for sustaining high compliance and improving stroke patient outcomes. SPANISH ABSTRACT: http://links.lww.com/IJEBH/A324.
- MeSH
- Time Factors MeSH
- Stroke * diagnostic imaging diagnosis MeSH
- Guideline Adherence MeSH
- Adult MeSH
- Evidence-Based Practice MeSH
- Humans MeSH
- Neuroimaging * methods standards MeSH
- Tomography, X-Ray Computed MeSH
- Practice Guidelines as Topic MeSH
- Check Tag
- Adult MeSH
- Humans MeSH
- Publication type
- Journal Article MeSH
- Geographicals
- United Kingdom MeSH
Radiologists utilize pictures from X-rays, magnetic resonance imaging, or computed tomography scans to diagnose bone cancer. Manual methods are labor-intensive and may need specialized knowledge. As a result, creating an automated process for distinguishing between malignant and healthy bone is essential. Bones that have cancer have a different texture than bones in unaffected areas. Diagnosing hematological illnesses relies on correct labeling and categorizing nucleated cells in the bone marrow. However, timely diagnosis and treatment are hampered by pathologists' need to identify specimens, which can be sensitive and time-consuming manually. Humanity's ability to evaluate and identify these more complicated illnesses has significantly been bolstered by the development of artificial intelligence, particularly machine, and deep learning. Conversely, much research and development is needed to enhance cancer cell identification-and lower false alarm rates. We built a deep learning model for morphological analysis to solve this problem. This paper introduces a novel deep convolutional neural network architecture in which hybrid multi-objective and category-based optimization algorithms are used to optimize the hyperparameters adaptively. Using the processed cell pictures as input, the proposed model is then trained with an optimized attention-based multi-scale convolutional neural network to identify the kind of cancer cells in the bone marrow. Extensive experiments are run on publicly available datasets, with the results being measured and evaluated using a wide range of performance indicators. In contrast to deep learning models that have already been trained, the total accuracy of 99.7% was determined to be superior.
- MeSH
- Algorithms MeSH
- Deep Learning * MeSH
- Bone Marrow diagnostic imaging pathology MeSH
- Humans MeSH
- Bone Neoplasms pathology diagnostic imaging diagnosis MeSH
- Neural Networks, Computer * MeSH
- Image Processing, Computer-Assisted methods MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH