Diagnostics

32 pages, 6508 KB

Open AccessArticle

An Explainable Web-Based Diagnostic System for Alzheimer’s Disease Using XRAI and Deep Learning on Brain MRI

by Serra Aksoy and Arij Daou

Diagnostics 2025, 15(20), 2559; https://doi.org/10.3390/diagnostics15202559 - 10 Oct 2025

Background: Alzheimer’s disease (AD) is a progressive neurodegenerative condition marked by cognitive decline and memory loss. Despite advancements in AI-driven neuroimaging analysis for AD detection, clinical deployment remains limited due to challenges in model interpretability and usability. Explainable AI (XAI) frameworks such as [...] Read more.

Background: Alzheimer’s disease (AD) is a progressive neurodegenerative condition marked by cognitive decline and memory loss. Despite advancements in AI-driven neuroimaging analysis for AD detection, clinical deployment remains limited due to challenges in model interpretability and usability. Explainable AI (XAI) frameworks such as XRAI offer potential to bridge this gap by providing clinically meaningful visualizations of model decision-making. Methods: This study developed a comprehensive, clinically deployable AI system for AD severity classification using 2D brain MRI data. Three deep learning architectures MobileNet-V3 Large, EfficientNet-B4, and ResNet-50 were trained on an augmented Kaggle dataset (33,984 images across four AD severity classes). The models were evaluated on both augmented and original datasets, with integrated XRAI explainability providing region-based attribution maps. A web-based clinical interface was built using Gradio to deliver real-time predictions and visual explanations. Results: MobileNet-V3 achieved the highest accuracy (99.18% on the augmented test set; 99.47% on the original dataset), while using the fewest parameters (4.2 M), confirming its efficiency and suitability for clinical use. XRAI visualizations aligned with known neuroanatomical patterns of AD progression, enhancing clinical interpretability. The web interface delivered sub-20 s inference with high classification confidence across all AD severity levels, successfully supporting real-world diagnostic workflows. Conclusions: This research presents the first systematic integration of XRAI into AD severity classification using MRI and deep learning. The MobileNet-V3-based system offers high accuracy, computational efficiency, and interpretability through a user-friendly clinical interface. These contributions demonstrate a practical pathway toward real-world adoption of explainable AI for early and accurate Alzheimer’s disease detection. Full article

(This article belongs to the Special Issue Alzheimer's Disease Diagnosis Based on Deep Learning)

► Show Figures

Figure 1

20 pages, 4773 KB

Open AccessArticle

Progressive Disease Image Generation with Ordinal-Aware Diffusion Models

by Meryem Mine Kurt, Ümit Mert Çağlar and Alptekin Temizel

Diagnostics 2025, 15(20), 2558; https://doi.org/10.3390/diagnostics15202558 - 10 Oct 2025

Abstract

Background/Objectives: Ulcerative Colitis (UC) lacks longitudinal visual data, which limits both disease progression modeling and the effectiveness of computer-aided diagnosis systems. These systems are further constrained by sparse intermediate disease stages and the discrete nature of the Mayo Endoscopic Score (MES). Meanwhile, synthetic [...] Read more.

Background/Objectives: Ulcerative Colitis (UC) lacks longitudinal visual data, which limits both disease progression modeling and the effectiveness of computer-aided diagnosis systems. These systems are further constrained by sparse intermediate disease stages and the discrete nature of the Mayo Endoscopic Score (MES). Meanwhile, synthetic image generation has made significant advances. In this paper, we propose novel ordinal embedding architectures for conditional diffusion models to generate realistic UC progression sequences from cross-sectional endoscopic images. Methods: By adapting Stable Diffusion v1.4 with two specialized ordinal embeddings (Basic Ordinal Embedder using linear interpolation and Additive Ordinal Embedder modeling cumulative pathological features), our framework converts discrete MES categories into continuous progression representations. Results: The Additive Ordinal Embedder outperforms alternatives, achieving superior distributional alignment (CMMD 0.4137, recall 0.6331) and disease consistency comparable to real data (Quadratic Weighted Kappa 0.8425, UMAP Silhouette Score 0.0571). The generated sequences exhibit smooth transitions between severity levels while maintaining anatomical fidelity. Conclusions: This work establishes a foundation for transforming static medical datasets into dynamic progression models and demonstrates that ordinal-aware embeddings can effectively capture disease severity relationships, enabling synthesis of underrepresented intermediate stages. These advances support applications in medical education, diagnosis, and synthetic data generation. Full article

(This article belongs to the Special Issue Computer-Aided Diagnosis in Endoscopy 2025)

► Show Figures

Figure 1

20 pages, 5042 KB

Open AccessArticle

Assessment of the Relationship Between Haller Cells, Accessory Maxillary Ostium, and Maxillary Sinus Pathologies: A Cross-Sectional CBCT Study

by İsmail Çapar, Çiğdem Şeker and Orhan Cicek

Diagnostics 2025, 15(20), 2557; https://doi.org/10.3390/diagnostics15202557 - 10 Oct 2025

Abstract

Background/Objectives: Sinonasal anatomical variations, particularly Haller cells (HCs) and the accessory maxillary ostium (AMO), are critical structural factors that may increase surgical risks in dental and otorhinolaryngological (ENT) procedures and predispose individuals to chronic sinusitis. This study aimed to investigate the relationship [...] Read more.

Background/Objectives: Sinonasal anatomical variations, particularly Haller cells (HCs) and the accessory maxillary ostium (AMO), are critical structural factors that may increase surgical risks in dental and otorhinolaryngological (ENT) procedures and predispose individuals to chronic sinusitis. This study aimed to investigate the relationship between HCs, AMO dimensions, maxillary sinus ostium, and sinus pathologies using cone-beam computed tomography (CBCT). Methods: In this cross-sectional retrospective study, CBCT images of 443 patients (226 males, mean age 48.4 ± 15.4 years; 217 females, mean age 46.1 ± 15.2 years) were analyzed. The presence of HCs, AMO, ostium narrowing, and ostium obstruction were recorded, along with ostium dimensions. Relationships between these variations and sinus pathologies were statistically evaluated, with a p-value < 0.05 considered significant. Results: HC prevalence was 34.5% on the right and 39.5% on the left, while AMO was present in 39.5% on the right and 34.5% on the left. Bilateral AMO was significantly associated with localized mucosal thickening, and partial opacification was more common in cases with ostium obstruction. Significant relationships were observed between HC presence and ostium narrowing. While HCs and ostium narrowing did not significantly influence maxillary sinus pathologies, sex (right OR = 0.335; left OR = 0.384; p < 0.001) and the AMO (right OR = 1.698, p = 0.018; left OR = 1.713, p = 0.014) were found to have a significant impact. Conclusions: It was concluded that (i) HCs may contribute to ostium narrowing and impaired sinus drainage, thereby increasing the risk of chronic sinusitis; (ii) the presence of a bilateral AMO is strongly associated with localized mucosal thickening; (iii) sex and the presence of an AMO emerge as independent predictors of maxillary sinus pathologies; and (iv) the careful evaluation of these anatomical variations using CBCT can support multidisciplinary treatment planning in both dental and ENT practice, enhance surgical safety, and help minimize postoperative complications. Full article

(This article belongs to the Section Medical Imaging and Theranostics)

13 pages, 1930 KB

Open AccessArticle

Peripheral Nerve Ultrasound Findings in Hereditary Transthyretin Amyloidosis in Brazil

by Antonio Edvan Camelo-Filho, Anna Paula Paranhos Miranda Covaleski, Lara Albuquerque Brito, Cleonisio Leite Rodrigues and Ana Lucila Moreira

Diagnostics 2025, 15(20), 2556; https://doi.org/10.3390/diagnostics15202556 - 10 Oct 2025

Abstract

Background/Objectives: Hereditary transthyretin amyloidosis (ATTRv) is an autosomal dominant disorder characterized by systemic deposition of amyloid fibrils, leading to peripheral neuropathy and multisystemic involvement. Peripheral nerve ultrasound is a promising tool for detecting structural nerve changes, yet its use in Latin American [...] Read more.

Background/Objectives: Hereditary transthyretin amyloidosis (ATTRv) is an autosomal dominant disorder characterized by systemic deposition of amyloid fibrils, leading to peripheral neuropathy and multisystemic involvement. Peripheral nerve ultrasound is a promising tool for detecting structural nerve changes, yet its use in Latin American populations is limited. This study aimed to characterize nerve ultrasound findings in Brazilian patients with ATTRv. Methods: We conducted a cross-sectional study of 72 genetically confirmed ATTRv individuals from two Brazilian centers. Participants were classified into symptomatic patients with polyneuropathy (n = 31) and asymptomatic TTR variant carriers (n = 41). All participants underwent a standardized neurological examination, and nerve ultrasound was used to assess the median nerve, brachial plexus, and C6 root. Cross-sectional areas (CSAs) from the right side were used for analysis and compared to reference values. Conclusions: Symptomatic patients showed increased CSAs in the median nerve (wrist: 10.17 mm², arm: 9.8 mm²), C6 root (8.55 mm²), and brachial plexus (70.82 mm²; all p < 0.05), but not in the forearm. Notably, asymptomatic carriers exhibited nerve enlargement in the median nerve at the wrist, the C6 root, and the brachial plexus, despite lacking clinical signs of neuropathy. Peripheral nerve enlargement in ATTRv affects both symptomatic patients and asymptomatic carriers, with a predilection for proximal and entrapment sites. These findings support the utility of nerve ultrasound as a non-invasive biomarker for early nerve involvement in ATTRv. Further studies are warranted to validate its role in disease monitoring and guide therapeutic interventions, especially in genetically at-risk populations. Full article

(This article belongs to the Special Issue Clinical Anatomy and Diagnosis of Peripheral Nervous System—2nd Edition)

► Show Figures

Figure 1

31 pages, 2953 KB

Open AccessArticle

A Balanced Multimodal Multi-Task Deep Learning Framework for Robust Patient-Specific Quality Assurance

by Xiaoyang Zeng, Awais Ahmed and Muhammad Hanif Tunio

Diagnostics 2025, 15(20), 2555; https://doi.org/10.3390/diagnostics15202555 - 10 Oct 2025

Abstract

Background: Multimodal Deep learning has emerged as a crucial method for automated patient-specific quality assurance (PSQA) in radiotherapy research. Integrating image-based dose matrices with tabular plan complexity metrics enables more accurate prediction of quality indicators, including the Gamma Passing Rate (GPR) and dose [...] Read more.

Background: Multimodal Deep learning has emerged as a crucial method for automated patient-specific quality assurance (PSQA) in radiotherapy research. Integrating image-based dose matrices with tabular plan complexity metrics enables more accurate prediction of quality indicators, including the Gamma Passing Rate (GPR) and dose difference (DD). However, modality imbalance remains a significant challenge, as tabular encoders often dominate training, suppressing image encoders and reducing model robustness. This issue becomes more pronounced under task heterogeneity, with GPR prediction relying more on tabular data, whereas dose difference prediction (DDP) depends heavily on image features. Methods: We propose BMMQA (Balanced Multi-modal Quality Assurance), a novel framework that achieves modality balance by adjusting modality-specific loss factors to control convergence dynamics. The framework introduces four key innovations: (1) task-specific fusion strategies (softmax-weighted attention for GPR regression and spatial cascading for DD prediction); (2) a balancing mechanism supported by Shapley values to quantify modality contributions; (3) a fast network forward mechanism for efficient computation of different modality combinations; and (4) a modality-contribution-based task weighting scheme for multi-task multimodal learning. A large-scale multimodal dataset comprising 1370 IMRT plans was curated in collaboration with Peking Union Medical College Hospital (PUMCH). Results: Experimental results demonstrate that, under the standard 2%/3 mm GPR criterion, BMMQA outperforms existing fusion baselines. Under the stricter 2%/2 mm criterion, it achieves a 15.7% reduction in mean absolute error (MAE). The framework also enhances robustness in critical failure cases (GPR < 90%) and achieves a peak SSIM of 0.964 in dose distribution prediction. Conclusions: Explicit modality balancing improves predictive accuracy and strengthens clinical trustworthiness by mitigating overreliance on a single modality. This work highlights the importance of addressing modality imbalance for building trustworthy and robust AI systems in PSQA and establishes a pioneering framework for multi-task multimodal learning. Full article

(This article belongs to the Special Issue Deep Learning in Medical and Biomedical Image Processing)

► Show Figures

Figure 1

13 pages, 1306 KB

Open AccessArticle

HMGB1 and Kallistatin: Novel Serological Markers for Differentiating Peritonsillar Cellulitis and Abscess

by Kadir Sinasi Bulut, Fatih Gul, Tuba Saadet Deveci Bulut, Burak Celik, Serkan Serifler and Mehmet Ali Babademez

Diagnostics 2025, 15(20), 2554; https://doi.org/10.3390/diagnostics15202554 - 10 Oct 2025

Abstract

Background/Objectives: Peritonsillar abscess (PTA) and cellulitis (PTC) often present with similar clinical features, making differentiation challenging despite imaging. This study evaluates the diagnostic performance of serum HMGB1 and kallistatin levels as potential independent biomarkers to distinguish PTA from PTC. Methods: In [...] Read more.

Background/Objectives: Peritonsillar abscess (PTA) and cellulitis (PTC) often present with similar clinical features, making differentiation challenging despite imaging. This study evaluates the diagnostic performance of serum HMGB1 and kallistatin levels as potential independent biomarkers to distinguish PTA from PTC. Methods: In this single-center prospective cohort study, 97 patients aged 18 to 65 years who met the inclusion criteria and presented with peritonsillar infection (39 PTA; 58 PTC) between February and July 2025 were enrolled. Serum levels of HMGB1, kallistatin, and routine inflammatory markers were measured and compared. Univariate and multivariate logistic regression analyses identified independent predictors for distinguishing PTA from PTC. Receiver operating characteristic (ROC) curve analysis assessed the diagnostic accuracy of biomarkers. Decision curve analysis (DCA) was performed to evaluate the clinical net benefit of individual biomarkers and their combinations across a range of threshold probabilities. Results: Compared to controls, patients with peritonsillar infection had significantly higher WBC, neutrophil, CRP, procalcitonin, and HMGB1 levels and significantly lower kallistatin levels (all p < 0.05). Within the infection group, PTA patients showed significantly higher CRP (p = 0.036) and HMGB1 (p = 0.003) levels and lower kallistatin (p < 0.001) levels compared to PTC patients. In univariate analysis, CRP, HMGB1, and kallistatin were significantly associated with PTA; however, in multivariate analysis, only elevated HMGB1 (OR: 1.21; 95% CI: 1.09–1.35; p < 0.001) and reduced kallistatin (OR: 0.395; 95% CI: 0.24–0.648; p < 0.001) remained independent predictors. ROC analysis showed that both HMGB1 and kallistatin demonstrated good discriminative ability in distinguishing PTA from PTC. DCA revealed that the three-biomarker combination (kallistatin + HMGB1 + CRP) achieved the highest mean net benefit (0.183) across all threshold probabilities, outperforming individual biomarkers (kallistatin: 0.131, HMGB1: 0.111, CRP: 0.099) and the two-biomarker model (0.176). The combined model maintained superior net benefit across threshold probabilities of 25–75%, indicating optimal clinical utility within this decision range. Conclusions: Serum HMGB1 and kallistatin may be effective adjunctive biomarkers for differentiating PTA from PTC. Full article

(This article belongs to the Section Clinical Diagnosis and Prognosis)

► Show Figures

Figure 1

14 pages, 714 KB

Open AccessArticle

Is Digital Maxillary Model Scanning Reliable in Individuals with Unilateral Cleft Lip and Palate?

by Elif Merve Mavi, Ozge Uslu-Akcam and Mehmet Okan Akcam

Diagnostics 2025, 15(20), 2553; https://doi.org/10.3390/diagnostics15202553 (registering DOI) - 10 Oct 2025

Abstract

Objective: To evaluate the measurements made on digital scans of maxillary plaster models in comparison with those obtained directly with a digital caliper on plaster models obtained from individuals with unilateral cleft lip and palate. Methods: This study included 42 unilateral cleft lip [...] Read more.

Objective: To evaluate the measurements made on digital scans of maxillary plaster models in comparison with those obtained directly with a digital caliper on plaster models obtained from individuals with unilateral cleft lip and palate. Methods: This study included 42 unilateral cleft lip and palate cases and a control group of 43 Angle Class I cases. The research material consisted of maxillary orthodontic plaster models obtained from these individuals and three-dimensional digital models obtained by scanning these models with a 3 Shape Trios scanner. A total of 12 anatomic reference points were used and six transverse dimension parameters were measured. The differences between the two groups were examined with a Student’s t-test. Intraclass correlation coefficients were calculated for repeatability and similarity evaluations. Results: Significant differences were found between the CLP and control groups for all parameters, with smaller values obtained in the CLP group. In the CLP group, when comparing the asymmetry of the right and left regions in the 3 Shape model, significant differences were observed regarding all parameters (p < 0.05); furthermore, there was a significant difference between the CLP and control groups (p < 0.05) in the asymmetry comparison. In both groups, there was no statistically significant difference in the measured parameters between the 3 Shape and digital caliper measurements. Conclusions: The measurements obtained after scanning plaster models from CLP individuals with the 3 Shape digital scanner are acceptable and reliable. It can be concluded that the transfer of CLP patients’ archived plaster models to the digital environment is reliable regarding scientific research and clinical measurements. Full article

(This article belongs to the Special Issue New Possibilities for Digital Diagnosis and Planning in Dentistry)

► Show Figures

Figure 1

2 pages, 143 KB

Open AccessReply

Reply to Wouter, K. Comment on “Lee et al. Influence of Aortoiliac Geometry on Non-Occlusive Thrombotic Risk Following Endovascular Repair of Abdominal Aortic Aneurysms. Diagnostics 2025, 15, 2134”

by Jeong In Lee, Dac Hong An Ngo, Hong Pil Hwang, Young Min Han and Hyo Sung Kwak

Diagnostics 2025, 15(20), 2552; https://doi.org/10.3390/diagnostics15202552 - 10 Oct 2025

Abstract

We would like to thank Dr [...] Full article

(This article belongs to the Section Medical Imaging and Theranostics)

2 pages, 140 KB

Open AccessComment

Comparing Iliac Artery Sizes to Explain Post-EVAR Non-Obstructive Thrombosis. Comment on Lee et al. Influence of Aortoiliac Geometry on Non-Occlusive Thrombotic Risk Following Endovascular Repair of Abdominal Aortic Aneurysms. Diagnostics 2025, 15, 2134

by Wouter Kok

Diagnostics 2025, 15(20), 2551; https://doi.org/10.3390/diagnostics15202551 - 10 Oct 2025

Abstract

In the paper by Lee et al [...] Full article

(This article belongs to the Special Issue Recent Advances in Diagnostic and Interventional Radiology)

16 pages, 456 KB

Open AccessReview

Forensic Odontology in the Digital Era: A Narrative Review of Current Methods and Emerging Trends

by Carmen Corina Radu, Timur Hogea, Cosmin Carașca and Casandra-Maria Radu

Diagnostics 2025, 15(20), 2550; https://doi.org/10.3390/diagnostics15202550 - 10 Oct 2025

Abstract

Background/Objectives: Forensic dental determination plays a central role in human identification, age estimation, and trauma analysis in medico-legal contexts. Traditional approaches—including clinical examination, odontometric analysis, and radiographic comparison—remain essential but are constrained by examiner subjectivity, population variability, and reduced applicability in fragmented or [...] Read more.

Background/Objectives: Forensic dental determination plays a central role in human identification, age estimation, and trauma analysis in medico-legal contexts. Traditional approaches—including clinical examination, odontometric analysis, and radiographic comparison—remain essential but are constrained by examiner subjectivity, population variability, and reduced applicability in fragmented or degraded remains. Recent advances in cone-beam computed tomography (CBCT), three-dimensional surface scanning, intraoral imaging, and artificial intelligence (AI) offer promising opportunities to enhance accuracy, reproducibility, and integration with multidisciplinary forensic evidence. The aim of this review is to synthesize conventional and emerging approaches in forensic odontology, critically evaluate their strengths and limitations, and highlight areas requiring validation. Methods: A structured literature search was performed in PubMed, Scopus, Web of Science, and Google Scholar for studies published between 2015 and 2025. Search terms combined forensic odontology, dental identification, CBCT, 3D scanning, intraoral imaging, and AI methodologies. From 108 records identified, 81 peer-reviewed articles met eligibility criteria and were included for analysis. Results: Digital methods such as CBCT, 3D scanning, and intraoral imaging demonstrated improved diagnostic consistency compared with conventional techniques. AI-driven tools—including automated age and sex estimation, bite mark analysis, and restorative pattern recognition—showed potential to enhance objectivity and efficiency, particularly in disaster victim identification. Persistent challenges include methodological heterogeneity, limited dataset diversity, ethical concerns, and issues of legal admissibility. Conclusions: Digital and AI-based approaches should complement, not replace, the expertise of forensic odontologists. Standardization, validation across diverse populations, ethical safeguards, and supportive legal frameworks are necessary to ensure global reliability and medico-legal applicability. Full article

(This article belongs to the Special Issue Advances in Dental Imaging, Oral Diagnosis, and Forensic Dentistry)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Diagnostics, Volume 15, Issue 20 (October-2 2025) – 10 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI