The impact of transfer learning on 3D deep learning convolutional neural network segmentation of the hippocampus in mild cognitive impairment and Alzheimer disease subjects

Recent Advancements in Neuroimaging-Based Alzheimer's Disease Prediction Using Deep Learning Approaches in e-Health: A Systematic Review

Purpose

Alzheimer's disease (AD) is a severe neurological disease that significantly impairs brain function. Timely identification of AD is essential for appropriate treatment and care. This comprehensive review intends to examine current developments in deep learning (DL) approaches with neuroimaging for AD diagnosis, where popular imaging types, reviews well-known online accessible data sets, and describes different algorithms used in DL for the correct initial evaluation of AD are presented.

Significance

Conventional diagnostic techniques, including medical evaluations and cognitive assessments, usually not identify the initial stages of Alzheimer's. Neuroimaging methods, when integrated with DL techniques, have demonstrated considerable potential in enhancing the diagnosis and categorization of AD. DL models have received significant interest due to their capability to identify AD in its early phases automatically, which reduces the mortality rate and treatment cost of AD.

Method

An extensive literature search was performed in leading scientific databases, concentrating on papers published from 2021 to 2025. Research leveraging DL models on different neuroimaging techniques such as magnetic resonance imaging (MRI), positron emission tomography, and functional magnetic resonance imaging (fMRI), and so forth. The review complies with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Results

Current developments show that CNN-based techniques, especially those utilizing hybrid and transfer learning frameworks, outperform conventional DL methods. Research employing the combination of multimodal neuroimaging data has demonstrated enhanced diagnostic precision. Still, challenges such as method interpretability, data heterogeneity, and limited data exist as significant issues.

Conclusion

DL has considerably improved the accuracy and reliability of AD diagnosis with neuroimaging. Regardless of issues with data accessibility and adaptability, current studies into the interpretability of models and multimodal fusion provide potential for clinical application. Further research should concentrate on standardized data sets, rigorous validation architectures, and understandable AI methodologies to enhance the effectiveness of DL methods in AD prediction.

The emerging role of artificial intelligence in neuropathology: Where are we and where do we want to go?

The field of neuropathology, a subspecialty of pathology which studies the diseases affecting the nervous system, is experiencing significant changes due to advancements in artificial intelligence (AI). Traditionally reliant on histological methods and clinical correlations, neuropathology is now experiencing a revolution due to the development of AI technologies like machine learning (ML) and deep learning (DL). These technologies enhance diagnostic accuracy, optimize workflows, and enable personalized treatment strategies. AI algorithms excel at analyzing histopathological images, often revealing subtle morphological changes missed by conventional methods. For example, deep learning models applied to digital pathology can effectively differentiate tumor grades and detect rare pathologies, leading to earlier and more precise diagnoses. Progress in neuroimaging is another helpful tool of AI, as enhanced analysis of MRI and CT scans supports early detection of neurodegenerative diseases. By identifying biomarkers and progression patterns, AI aids in timely therapeutic interventions, potentially slowing disease progression. In molecular pathology, AI's ability to analyze complex genomic data helps uncover the genetic and molecular basis of neuropathological conditions, facilitating personalized treatment plans. AI-driven automation streamlines routine diagnostic tasks, allowing pathologists to focus on complex cases, especially in settings with limited resources. This review explores AI's integration into neuropathology, highlighting its current applications, benefits, challenges, and future directions.

Deep learning techniques for Alzheimer's disease detection in 3D imaging: A systematic review

Background and Aims

Alzheimer's disease (AD) is a degenerative neurological condition that worsens over time and leads to deterioration in cognitive abilities, reduced memory, and, eventually, a decrease in overall functioning. Timely and correct identification of Alzheimer's is essential for effective treatment. The systematic study specifically examines the application of deep learning (DL) algorithms in identifying AD using three-dimensional (3D) imaging methods. The main goal is to evaluate these methods' current state, efficiency, and potential enhancements, offering valuable insights into how DL could improve AD's rapid and precise diagnosis.

Methods

We searched different online repositories, such as IEEE Xplore, Elsevier, MDPI, PubMed Central, Science Direct, ACM, Springer, and others, to thoroughly summarize current research on DL methods to diagnose AD by analyzing 3D imaging data published between 2020 and 2024. We use PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to ensure the organization and understandability of the information collection process. We thoroughly analyzed the literature to determine the primary techniques used in these investigations and their findings.

Results and Conclusion

The ability of convolutional neural networks (CNNs) and their variations, including 3D CNNs and recurrent neural networks, to detect both temporal and spatial characteristics in volumetric data has led to their widespread use. Methods such as transfer learning, combining multimodal data, and using attention procedures have improved models' precision and reliability. We selected 87 articles for evaluation. Out of these, 31 papers included various concepts, explanations, and elucidations of models and theories, while the other 56 papers primarily concentrated on issues related to practical implementation. This article introduces popular imaging types, 3D imaging for Alzheimer's detection, discusses the benefits and restrictions of the DL-based approach to AD assessment, and gives a view toward future developments resulting from critical evaluation.

Revolutionizing anemia detection: integrative machine learning models and advanced attention mechanisms

This study addresses the critical issue of anemia detection using machine learning (ML) techniques. Although a widespread blood disorder with significant health implications, anemia often remains undetected. This necessitates timely and efficient diagnostic methods, as traditional approaches that rely on manual assessment are time-consuming and subjective. The present study explored the application of ML - particularly classification models, such as logistic regression, decision trees, random forest, support vector machines, Naïve Bayes, and k-nearest neighbors - in conjunction with innovative models incorporating attention modules and spatial attention to detect anemia. The proposed models demonstrated promising results, achieving high accuracy, precision, recall, and F1 scores for both textual and image datasets. In addition, an integrated approach that combines textual and image data was found to outperform the individual modalities. Specifically, the proposed AlexNet Multiple Spatial Attention model achieved an exceptional accuracy of 99.58%, emphasizing its potential to revolutionize automated anemia detection. The results of ablation studies confirm the significance of key components - including the blue-green-red, multiple, and spatial attentions - in enhancing model performance. Overall, this study presents a comprehensive and innovative framework for noninvasive anemia detection, contributing valuable insights to the field.

Transfer learning for anatomical structure segmentation in otorhinolaryngology microsurgery

BACKGROUND

Reducing the annotation burden is an active and meaningful area of artificial intelligence (AI) research.

METHODS

Multiple datasets for the segmentation of two landmarks were constructed based on 41 257 labelled images and 6 different microsurgical scenarios. These datasets were trained using the multi-stage transfer learning (TL) methodology.

RESULTS

The multi-stage TL enhanced segmentation performance over baseline (mIOU 0.6892 vs. 0.8869). Besides, Convolutional Neural Networks (CNNs) achieved a robust performance (mIOU 0.8917 vs. 0.8603) even when the training dataset size was reduced from 90% (30 078 images) to 10% (3342 images). When directly applying the weight from one certain surgical scenario to recognise the same target in images of other scenarios without training, CNNs still obtained an optimal mIOU of 0.6190 ± 0.0789.

CONCLUSIONS

Model performance can be improved with TL in datasets with reduced size and increased complexity. It is feasible for data-based domain adaptation among different microsurgical fields.

AI in imaging: the regulatory landscape

Artificial intelligence (AI) methods have been applied to medical imaging for several decades, but in the last few years, the number of publications and the number of AI-enabled medical devices coming on the market have significantly increased. While some AI-enabled approaches are proving very valuable, systematic reviews of the AI imaging field identify significant weaknesses in a significant proportion of the literature. Medical device regulators have recently become more proactive in publishing guidance documents and recognizing standards that will require that the development and validation of AI-enabled medical devices need to be more rigorous than required for tradition "rule-based" software. In particular, developers are required to better identify and mitigate risks (such as bias) that arise in AI-enabled devices, and to ensure that the devices are validated in a realistic clinical setting to ensure their output is clinically meaningful. While this evolving regulatory landscape will mean that device developers will take longer to bring novel AI-based medical imaging devices to market, such additional rigour is necessary to address existing weaknesses in the field and ensure that patients and healthcare professionals can trust AI-enabled devices. There would also be benefits in the academic community taking into account this regulatory framework, to improve the quality of the literature and make it easier for academically developed AI tools to make the transition to medical devices that impact healthcare.

Toward hippocampal volume measures on ultra-high field magnetic resonance imaging: a comprehensive comparison study between deep learning and conventional approaches

The hippocampus is a complex brain structure that plays an important role in various cognitive aspects such as memory, intelligence, executive function, and path integration. The volume of this highly plastic structure is identified as one of the most important biomarkers of specific neuropsychiatric and neurodegenerative diseases. It has also been extensively investigated in numerous aging studies. However, recent studies on aging show that the performance of conventional approaches in measuring the hippocampal volume is still far from satisfactory, especially in terms of delivering longitudinal measures from ultra-high field magnetic resonance images (MRIs), which can visualize more boundary details. The advancement of deep learning provides an alternative solution to measuring the hippocampal volume. In this work, we comprehensively compared a deep learning pipeline based on nnU-Net with several conventional approaches including Freesurfer, FSL and DARTEL, for automatically delivering hippocampal volumes: (1) Firstly, we evaluated the segmentation accuracy and precision on a public dataset through cross-validation. Results showed that the deep learning pipeline had the lowest mean (L = 1.5%, R = 1.7%) and the lowest standard deviation (L = 5.2%, R = 6.2%) in terms of volume percentage error. (2) Secondly, sub-millimeter MRIs of a group of healthy adults with test-retest 3T and 7T sessions were used to extensively assess the test-retest reliability. Results showed that the deep learning pipeline achieved very high intraclass correlation coefficients (L = 0.990, R = 0.986 for 7T; L = 0.985, R = 0.983 for 3T) and very small volume percentage differences (L = 1.2%, R = 0.9% for 7T; L = 1.3%, R = 1.3% for 3T). (3) Thirdly, a Bayesian linear mixed effect model was constructed with respect to the hippocampal volumes of two healthy adult datasets with longitudinal 7T scans and one disease-related longitudinal dataset. It was found that the deep learning pipeline detected both the subtle and disease-related changes over time with high sensitivity as well as the mild differences across subjects. Comparison results from the aforementioned three aspects showed that the deep learning pipeline significantly outperformed the conventional approaches by large margins. Results also showed that the deep learning pipeline can better accommodate longitudinal analysis purposes.

Machine Learning for Brain MRI Data Harmonisation: A Systematic Review

BACKGROUND

Magnetic Resonance Imaging (MRI) data collected from multiple centres can be heterogeneous due to factors such as the scanner used and the site location. To reduce this heterogeneity, the data needs to be harmonised. In recent years, machine learning (ML) has been used to solve different types of problems related to MRI data, showing great promise.

OBJECTIVE

This study explores how well various ML algorithms perform in harmonising MRI data, both implicitly and explicitly, by summarising the findings in relevant peer-reviewed articles. Furthermore, it provides guidelines for the use of current methods and identifies potential future research directions.

METHOD

This review covers articles published through PubMed, Web of Science, and IEEE databases through June 2022. Data from studies were analysed based on the criteria of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Quality assessment questions were derived to assess the quality of the included publications.

RESULTS

a total of 41 articles published between 2015 and 2022 were identified and analysed. In the review, MRI data has been found to be harmonised either in an implicit (n = 21) or an explicit (n = 20) way. Three MRI modalities were identified: structural MRI (n = 28), diffusion MRI (n = 7) and functional MRI (n = 6).

CONCLUSION

Various ML techniques have been employed to harmonise different types of MRI data. There is currently a lack of consistent evaluation methods and metrics used across studies, and it is recommended that the issue be addressed in future studies. Harmonisation of MRI data using ML shows promises in improving performance for ML downstream tasks, while caution should be exercised when using ML-harmonised data for direct interpretation.

Adapting UK Biobank imaging for use in a routine memory clinic setting: The Oxford Brain Health Clinic

The Oxford Brain Health Clinic (BHC) is a joint clinical-research service that provides memory clinic patients and clinicians access to high-quality assessments not routinely available, including brain MRI aligned with the UK Biobank imaging study (UKB). In this work we present how we 1) adapted the UKB MRI acquisition protocol to be suitable for memory clinic patients, 2) modified the imaging analysis pipeline to extract measures that are in line with radiology reports and 3) explored the alignment of measures from BHC patients to the largest brain MRI study in the world (ultimately 100,000 participants). Adaptations of the UKB acquisition protocol for BHC patients include dividing the scan into core and optional sequences (i.e., additional imaging modalities) to improve patients' tolerance for the MRI assessment. We adapted the UKB structural MRI analysis pipeline to take into account the characteristics of a memory clinic population (e.g., high amount of white matter hyperintensities and hippocampal atrophy). We then compared the imaging derived phenotypes (IDPs) extracted from the structural scans to visual ratings from radiology reports, non-imaging factors (age, cognition) and to reference distributions derived from UKB data. Of the first 108 BHC attendees (August 2020-November 2021), 92.5 % completed the clinical scans, 88.0 % consented to use of data for research, and 43.5 % completed the additional research sequences, demonstrating that the protocol is well tolerated. The high rates of consent to research makes this a valuable real-world quality research dataset routinely captured in a clinical service. Modified tissue-type segmentation with lesion masking greatly improved grey matter volume estimation. CSF-masking marginally improved hippocampal segmentation. The IDPs were in line with radiology reports and showed significant associations with age and cognitive performance, in line with the literature. Due to the age difference between memory clinic patients of the BHC (age range 65-101 years, average 78.3 years) and UKB participants (44-82 years, average 64 years), additional scans on elderly healthy controls are needed to improve reference distributions. Current and future work aims to integrate automated quantitative measures in the radiology reports and evaluate their clinical utility.

Deep learning in neuroimaging data analysis: Applications, challenges, and solutions

Methods for the analysis of neuroimaging data have advanced significantly since the beginning of neuroscience as a scientific discipline. Today, sophisticated statistical procedures allow us to examine complex multivariate patterns, however most of them are still constrained by assuming inherent linearity of neural processes. Here, we discuss a group of machine learning methods, called deep learning, which have drawn much attention in and outside the field of neuroscience in recent years and hold the potential to surpass the mentioned limitations. Firstly, we describe and explain the essential concepts in deep learning: the structure and the computational operations that allow deep models to learn. After that, we move to the most common applications of deep learning in neuroimaging data analysis: prediction of outcome, interpretation of internal representations, generation of synthetic data and segmentation. In the next section we present issues that deep learning poses, which concerns multidimensionality and multimodality of data, overfitting and computational cost, and propose possible solutions. Lastly, we discuss the current reach of DL usage in all the common applications in neuroimaging data analysis, where we consider the promise of multimodality, capability of processing raw data, and advanced visualization strategies. We identify research gaps, such as focusing on a limited number of criterion variables and the lack of a well-defined strategy for choosing architecture and hyperparameters. Furthermore, we talk about the possibility of conducting research with constructs that have been ignored so far or/and moving toward frameworks, such as RDoC, the potential of transfer learning and generation of synthetic data.