Search Results (95)

Background: Chagas disease, caused by Trypanosoma cruzi, remains endemic throughout Latin America but is increasingly reported in urban areas due to migration and vector adaptation. The cardiac form is the most severe manifestation, associated with arrhythmia, mural thrombus formation, and a high risk of cardioembolic events. Stroke secondary to Chagas cardiomyopathy is uncommon and poses diagnostic and therapeutic challenges. Case Presentation: A 58-year-old woman with serologic evidence of T. cruzi infection presented with sudden-onset dyspnea, oppressive chest pain, and left-sided weakness. Neurological examination revealed left brachiocrural hemiparesis and mild dysarthria (NIHSS = 9). Non-contrast cranial CT showed an acute infarct in the right middle cerebral artery territory (ASPECTS = 7). Electrocardiography demonstrated typical atrial flutter with variable conduction, and transthoracic echocardiography revealed a markedly dilated left atrium containing a mural thrombus and a left ventricular ejection fraction of 45%. Intravenous thrombolysis with alteplase (0.9 mg/kg) was administered within 4.5 h of symptom onset. Pharmacologic rhythm control was achieved using intravenous and oral amiodarone, followed by oral anticoagulation with warfarin (target INR 2.0–3.0) after excluding hemorrhagic transformation. The patient showed rapid neurological improvement (NIHSS reduction from 9 to 2) and was discharged on day 10 with minimal residual deficit (mRS = 1), sinus rhythm, and stable hemodynamics. Conclusions: This case highlights the rare coexistence of Chagas cardiomyopathy, atrial flutter, and cardioembolic stroke due to left atrial thrombus. Early recognition, adherence to evidence-based guidelines, and multidisciplinary management were key to achieving a favorable outcome. Timely diagnosis and intervention remain crucial to preventing severe complications in patients with Chagas disease. Full article

Background/Objectives: Non-dilated left ventricular cardiomyopathy (NDLVC) is a recently defined clinical entity associated with increased risk of ventricular arrhythmias (VA) and sudden cardiac death (SCD), despite preserved LV geometry. The role and temporal variability of noninvasive electrocardiographic (ECG) risk markers in this population remain insufficiently characterized. To assess the temporal variability of ECG-derived risk markers in patients with NDLVC and explore their association with major adverse cardiac events, including heart failure (HF) and VA hospitalization. Methods: We prospectively studied 55 patients with NDLVC who underwent cardiac magnetic resonance imaging and serial 24 h Holter monitoring, signal-averaged ECG, and standard 12-lead ECG over a one-year period. Patients were followed up for 39.5 ± 8.6 months. Nine ECG-based risk markers were analyzed, including premature ventricular contraction (PVC) burden, non-sustained ventricular tachycardia (NSVT) occurrence, its maximum rate and maximum beats, mean QTc interval, standard deviation of NN intervals (SDNN), deceleration capacity (DC), heart rate turbulence onset and slope (TO/TS), T-wave alternans (TWA), and late potentials. Clinical outcomes were HF and VA hospitalization. Logistic regression was used to evaluate associations between changes in ECG parameters and outcomes. Results: A change (from positive to negative and vice versa) in at least one ECG parameter was detected in 67.3% of patients, with the highest variability observed in TWA (34.5%), NSVT (23.6%), and PVC burden (23.6%). Despite this variability, only SDNN was significantly associated with increased risk of VA hospitalization during follow-up (OR = 0.98, 95% CI: 0.97–0.99, p = 0.006). No ECG changes were associated with HF hospitalization. Conclusions: Patients with NDLVC exhibit substantial temporal variability in noninvasive ECG risk markers. While most changes do not correlate with clinical events, an inverse association was found between SDNN and VA risk. These findings support the ongoing evaluation and the necessity to identify more effective risk stratification markers in this subgroup of patients. Full article

This study investigates FLNC mutations in Chinese cardiomyopathy patients. Background: Inherited cardiomyopathies, including dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM), restrictive cardiomyopathy (RCM), and arrhythmogenic right ventricular cardiomyopathy (ARVC) are major heart failure causes. FLNC, critical for muscle structure, is implicated in myofibrillar myopathy and isolated DCM (3–4% cases) with ventricular arrhythmias. Missense variants are linked to HCM and protein aggregation. A cohort of 25 patients with pathogenic/likely pathogenic FLNC mutations (2022–2025, Beijing Anzhen Hospital) underwent whole-exome sequencing (WES) using IDT kit 1.0/Hiseq 4000. Variants were classified via the American College of Medical Genetics and Genomics (ACMG) guidelines. Clinical data (echocardiography, CMR, labs) and follow-up data (prognosis, meds, and family history) were collected. The statistics used SPSS (p < 0.05). The mean age was 38 ± 14.6 years (13 males). There were 25 FLNC mutations: 12 single nucleotide polymorphisms (SNPs), 5 deletions, 2 duplications, and 3 deletion-insertions, classified as 6 pathogenic, 16 likely pathogenic, and 3 variants of uncertain significance (VUS). Diagnoses: 24% dilated cardiomyopathy (DCM), 8% hypertrophic cardiomyopathy (HCM), and 4% left ventricular non-compaction. Nonsense mutation carriers exhibited significantly higher tricuspid regurgitation prevalence compared to frameshift mutation carriers (6/9 vs. 2/10; p = 0.04). Echocardiography revealed reduced left ventricular ejection fraction (LVEF) (41.5 ± 14.1%), with statistically significant differences in fractional shortening (p = 0.024) and aortic root diameter (p = 0.028). Pedigree analysis confirmed that a frameshift mutation (LP) co-segregated with familial DCM and was associated with severe phenotypes, including sudden cardiac death. Furthermore, nonsense FLNC mutations correlated with increased tricuspid regurgitation severity, smaller aortic root dimensions, and reduced pulmonary artery flow velocity. Full article

Background: Left atrial strain (LAS) derived from speckle-tracking echocardiography (STE) provides a sensitive, load-dependent measure of atrial function and ventricular filling pressures. Data on LAS in paediatric cardiomyopathies are still scarce; therefore, this study aimed to assess LA phasic function in dilated (DCM) and hypertrophic (HCM) cardiomyopathy and to determine its relationship with clinical and echocardiographic indices of disease severity. Methods: We conducted a cross-sectional case–control study that included 84 children (DCM n = 29, HCM n = 29, control n = 26) who underwent comprehensive clinical and echocardiography evaluation, including LAS parameters (reservoir—LASr; conduit—LAScd; and contractile—LASct). Group comparisons were performed using ANOVA or Kruskal–Wallis tests with post hoc adjustments, and correlations were analysed using Pearson’s or Spearman’s coefficients. Multivariable linear and logistic regression models were adjusted for age, body surface area (BSA), heart rate (HR), and blood pressure (BP) percentiles. Results: LASr and LAScd were significantly reduced in both cardiomyopathy groups compared with controls (p < 0.001), following a graded pattern (DCM < HCM < control). In DCM, lower LASr was independently associated with higher left atrial volume index (LAVi) and elevated E/E′ ratio, whereas in HCM, septal hypertrophy (IVSd Z-score) and log NT-proBNP were dominant determinants of impaired LASr. In logistic regression, LASr (OR = 0.93, p = 0.016) and LAScd (OR = 1.21, p = 0.001) independently predicted severe NYHA/Ross functional class after covariate adjustment, while LASct showed no significant association. Conclusions: These findings demonstrate that LA reservoir and conduit strain are markedly impaired in paediatric cardiomyopathy and are strongly linked to structural remodelling and functional limitation, underscoring their value as sensitive non-invasive markers of disease severity. Full article

In 2018, the Food and Drug Administration (FDA) reported cases of dilated cardiomyopathy (DCM) in dogs, including breeds without known genetic predisposition, fed diets containing a high proportion of legumes or potatoes, many labeled grain-free. Despite concerns, grain-free diets remain increasingly popular. This narrative review focuses on the peer-reviewed literature and summarizes recent studies evaluating the effects of grain-free diets on canine cardiac health, with emphasis on taurine deficiency, metabolic alterations, and emerging hypotheses extending beyond nutritional inadequacy. These findings suggest a strong link between diet and DCM, particularly regarding legumes. Dogs of various breeds showed larger left ventricular diameters, reduced systolic function, and increased premature ventricular complexes when fed non-traditional, grain-free, legume-rich diets compared to those on traditional, low-legume diets. Many affected dogs improved clinically and functionally with dietary changes and treatment, suggesting nutritional DCM could be reversible. Though mechanisms remain unclear, the research highlights the potential roles of legumes, particularly peas, in the gut microbiota and fiber-related bile acid metabolism. This review also distinguishes taurine-deficiency-related DCM as a separate form, with certain breeds, notably Golden Retrievers, being more susceptible. Overall, further studies are required to better understand the role of nutrition in canine cardiac health. Full article

The transcriptional regulation of MYH6 and MYH7 genes has been extensively investigated in healthy versus failing hearts; however, their expression dynamics in healthy human hearts across age and sex, particularly in the context of cardiovascular risk factors such as hypertension, remain poorly characterised. This study aimed to carry out a reanalysis of MYH6 and MYH7 transcript levels in a large cohort of non-failing human left ventricular samples, stratified by sex, age, and hypertensive status. Furthermore, we examined how age and sex influence gene expression differences between non-failing and failing hearts, the latter affected by dilated cardiomyopathy (DCM). Normalised expression values for MYH6 and MYH7 transcripts from both healthy and failing left ventricles were extracted using the GEO2R online analysis tool from the publicly available RNA-sequencing library GSE141910. This library provides transcriptomic profiles of left ventricular (LV) tissue from both healthy individuals and patients with cardiomyopathies. The Mann–Whitney U test was employed for pairwise comparisons between different groups stratified by sex, age, and hypertensive status. Statistical analysis demonstrates sex-specific differences in MYH6 and MYH7 expression in healthy left ventricles, with postmenopausal females (aged > 50 years) with hypertension emerging as a distinct group. Conversely, in end-stage DCM hearts, the expression levels of both myosin genes seemed to be primarily influenced by disease-related pathophysiological mechanisms rather than by sex or age. Comparison between healthy and failing hearts revealed a consistent and significant downregulation of MYH6 in all comparisons, irrespective of sex or age. On the other hand, MYH7 expression exhibited greater variability, particularly among males, with age and hypertensive status influencing its expression. The results underscore the importance of considering age, sex, and comorbidities in interpreting cardiac gene expression patterns and highlight potential regulatory divergence in contractile gene expression during cardiac remodelling. Full article

Chronic alcohol misuse is the leading cause of non-ischemic dilated cardiomyopathy, and the molecular mechanisms underlying the development of alcohol-associated cardiomyopathy (ACM), particularly regarding sex-specific susceptibility and mitochondrial contributions, are not fully known. In this study, we utilized a preclinical model of chronic + binge ethanol consumption to investigate sex differences in disease severity and mitochondrial function. Male and female C57BL/6J mice were fed ethanol or control liquid diets for 30 days, with 2 binge episodes on days 10 and 30. Cardiac morphology was assessed via echocardiography and cardiac function via left ventricular pressure–volume catheterization. Mitochondrial function was evaluated ex vivo using Seahorse XF analysis, ATP luminescence, and AmplexTM Red fluorescence in isolated ventricular mitochondria. Ethanol feeding induced significant cardiac dysfunction and increased transcriptional expression of inflammatory and fibrotic markers in males, while these effects were not seen in females. Despite these sex-specific cardiac effects, mitochondrial respiration, ATP production, collagen protein expression, and oxidative stress were not significantly altered following alcohol exposure in either sex. Further investigation is warranted to assess the potential role of ovarian hormones in this female cardioprotection against chronic + binge ethanol. Full article

Background/Objectives: Fabry disease (FD) is a rare X-linked disease caused by the deficient activity of the enzyme α-galactosidase A. Cardiac involvement is particularly critical, often determining the disease prognosis. Epidemiological data on FD in Portugal are limited and inconsistent, highlighting the need for targeted screening. The F-CHECK study aimed to determine the prevalence of FD through the systematic screening of a Portuguese cohort of patients with unexplained cardiomyopathies. Methods: This multicenter observational study (NCT05409846) assessed the prevalence and clinical characteristics of FD in a Portuguese cohort (n = 409) of patients from 10 central hospitals who presented with unexplained cardiomyopathies, including idiopathic hypertrophic cardiomyopathy (HCM), left ventricular hypertrophy, dilated-phase HCM, and dilated cardiomyopathy with late gadolinium enhancement in the inferolateral segment. Screening was performed using dried blood spot assays to measure α-galactosidase A activity and/or by GLA gene sequencing in whole-blood samples. Results: FD was diagnosed in 14 patients, corresponding to a prevalence of 3.4%. FD diagnosis was significantly associated with systemic manifestations such as acroparesthesias (p = 0.027) and angiokeratomas (p = 0.003), as well as an increased risk of prior arrhythmic events (p = 0.021) and cerebrovascular disease (p = 0.016). Most FD patients (57%) presented a non-founder mutation in the GLA gene; however, they were pathogenically relevant. Conclusions: The observed 3.4% prevalence highlights the importance of systematic FD screening among Portuguese patients with unexplained cardiomyopathy, extending beyond classic hypertrophic presentations to dilated forms. Specific clinical signs, electrocardiogram findings, and cardiac imaging features can serve as valuable indicators to guide targeted genetic testing for FD. Full article

Left ventricular non-compaction cardiomyopathy (LVNC) is an uncommon myocardial phenotype characterized by prominent trabeculae and deep blood-filled recesses. The expanding use of cardiac magnetic resonance (CMR) has increased detection, yet uncertainty persists about whether LVNC is a distinct disease or a phenotype that overlaps with other cardiomyopathies. LVNC expression reflects the interplay among genotype, sex, ancestry, and hemodynamic load and thus serves as a model for precision cardiology. We conducted a narrative review of literature published between January 2000 and April 2025 in major databases. We included clinical studies with at least 10 patients, meta-analyses, reviews, and consensus statements addressing pathophysiology, genetics, diagnosis, prognosis, and treatment. Sarcomeric variants account for a substantial fraction of cases and connect LVNC with dilated and hypertrophic cardiomyopathies. Echocardiographic and CMR criteria identify the phenotype but blur the boundary between physiological and pathological hypertrabeculation. Fibrosis on late gadolinium enhancement and systolic dysfunction are consistently associated with worse outcomes. Current management largely adapts heart-failure strategies, including neurohormonal blockade, SGLT2 inhibitors, and implantable cardioverter-defibrillators in selected high-risk patients. Optimal care integrates clinical, imaging, and genetic information. The lack of universal diagnostic criteria highlights the need for prospective studies and consensus to standardize diagnosis and treatment. Future algorithms that combine multi-omics, quantitative imaging, and AI-based risk prediction could individualize surveillance, pharmacotherapy, and device therapy. Full article

Background/Objectives: Patients with non-valvular heart failure frequently develop valvular disease. However, the prevalence of valvular disease across patients with different heart failure etiologies remains underexplored. This study aimed to investigate the burden of VHD among patients with non-valvular heart failure, and secondly evaluate its association with cardiopulmonary test. Methods: We analyzed data from patients with non-valvular heart failure (HF) who were evaluated as outpatients at the HF clinic between February 2020 and November 2024. Patients were categorized into three groups: coronary artery disease-related HF (CAD-HF), dilated cardiomyopathy (DCM), and other causes (e.g., hypertension, diabetes, and various cardiomyopathies). Demographic and clinical characteristics, as well as echocardiographic and cardiopulmonary exercise testing (CPET) results, were evaluated. Results: Among all groups mild mitral regurgitation (MR) was the most common valvular disease, followed by mild tricuspid regurgitation (TR). Patients with CAD-HF frequently had mild aortic regurgitation (AR) compared to DCM (23.6% vs. 14.9%, p = 0.05). In the CPET subgroup, which included 41 patients who consented to participate, in patients with moderate-to-severe VHD had significantly lower VO₂/HR (oxygen pulse), VO₂max, and OUES, indicating worsened functional capacity despite similar left ventricular ejection fraction. Hypertension and atrial fibrillation were independently associated with greater valvular disease severity on multivariable analysis. Conclusions: No significant differences in valvular disease between patients with DCM and CAD-HF were documented, apart from a higher prevalence of mild AR in the CAD-HF group. Patients with moderate-to-severe valvular regurgitation demonstrated worse cardiopulmonary performance, regardless of ejection fraction, highlighting the important role of CPET in evaluating the functional impact of valvular heart disease in this population. Full article

Background: Non-ischemic dilated cardiomyopathy (DCM) is a heterogeneous myocardial disease associated with variable progression and an increased risk of major adverse cardiovascular events (MACEs). Cardiovascular magnetic resonance (CMR) allows the comprehensive evaluation of myocardial structure, function, and fibrosis. This prospective study aimed to assess the prognostic value of CMR-derived global longitudinal strain (GLS) and left ventricular (LV) torsion in patients with DCM. Methods: We prospectively enrolled 150 patients with newly diagnosed non-ischemic DCM and 100 age- and sex-matched healthy controls. All participants underwent standardized CMR protocols including cine imaging, late gadolinium enhancement (LGE), and feature-tracking analysis for myocardial deformation. LV volumes, ejection fraction (LVEF), GLS, and LV torsion were quantified. The primary endpoint was the first occurrence of MACE, defined as cardiac death, sustained ventricular arrhythmia, or heart failure hospitalization. The median follow-up was 33 months. Results: Compared to controls, DCM patients had significantly impaired LV function and myocardial mechanics: lower LVEF (35.1% vs. 65.2%, p < 0.001), reduced GLS (−9.2% vs. −19.7%, p < 0.001), and diminished LV torsion (1.04 vs. 1.95 °/cm, p < 0.001). GLS ≤ −8.6% was independently associated with increased MACE risk (adjusted hazard ratio [HR]: 1.09; 95% confidence interval [CI]: 1.01–1.61; p < 0.01). Similarly, reduced LV torsion predicted adverse events (adjusted HR: 1.37; 95% CI: 1.03–1.81; p < 0.01). The presence of LGE (42% of patients) further stratified risk (HR: 2.86; 95% CI: 1.48–12.52; p < 0.001). Conclusions: CMR-derived GLS and LV torsion are strong, independent predictors of adverse outcomes in DCM. Their integration into routine imaging protocols enhances risk stratification beyond conventional metrics such as LVEF and LGE. These findings support the use of myocardial deformation analysis in the comprehensive evaluation of patients with DCM. Full article

Background: The widespread use of cardiac magnetic resonance imaging (MRI) in clinical practice has enabled the identification of numerous patients with evident damage from previous myocarditis, whether known or unknown. For years, myocardial fibrosis has been a topic of interest due to its established correlation with arrhythmic events in various clinical settings, including ischemic heart disease, dilated cardiomyopathy, and hypertrophic cardiomyopathy. MIBG scintigraphy is a method widely used in patients who are candidates for defibrillator implantation or have experienced heart failure. This examination evaluates the sympathetic innervation of the myocardium. Objective: To assess the real arrhythmogenic risk of non-ischemic scars identified in symptomatic or asymptomatic patients through the use of MIBG. Methods: Patients were retrospectively selected based on the presence of non-ischemic myocardial fibrosis detected by cardiac MRI, consistent with a myocarditis outcome (even in the absence of a clear history of myocarditis). These patients underwent myocardial scintigraphy with MIBG using a tomographic technique. Results: A total of 50 patients (41 males, mean age 51 ± 16 years) who underwent MRI from 2019 to June 2024 were selected. The primary indication for MRI was ventricular ectopic extrasystoles detected on Holter ECG (n = 12, 54%), while five patients underwent MRI following a known acute infectious event (23%, including three cases of COVID-19 infection). All symptomatic patients presented with chest pain in the acute phase, accompanied by elevated hsTNI levels (mean value: 437 pg/mL). The MRI findings showed normal ventricular volumes (LV: 80 mL/m², RV: 81 mL/m²) and normal ejection fractions (56% and 53%, respectively). The mean native T1 mapping value was 1013 ms (normal range: 950–1050). T2 mapping values were altered in the 5 patients who underwent MRI during the acute phase (mean value: 57 ms), without segmentation. Additionally, three patients had non-tamponade pericardial effusion. All patients exhibited LGE (nine subepicardial, seven midwall, six patchy). All patients underwent myocardial scintigraphy with MIBG at least 6 months after the acute event, with only one case yielding a positive result. This patient, a 57-year-old male, had the most severe clinical presentation, including more than 65,000 premature ventricular beats (PVBs) and multiple episodes of paroxysmal supraventricular tachycardia (PSVT) recorded on Holter ECG. MRI findings showed severe left ventricular dysfunction, a slightly dilated LV, and midwall LGE at the septum, coinciding with hypokinetic areas. Conclusions: MIBG scintigraphy could be a useful tool in assessing arrhythmic risk in patients with previous myocarditis. It could help reduce the clinical burden of incidental findings of non-ischemic LGE, which does not appear to be independently associated with an increased risk profile. Full article

Background: Alternative splicing is an important mechanism of transcriptomic and proteomic diversity and is progressively involved in cardiovascular disease (CVD) pathogenesis. Serum response factor (SRF), a critical transcription factor in cardiac development and function, may itself undergo splicing regulation, potentially altering its function in disease states. Objective: The objective of this study is to identify SRF-associated alternative splicing events in cardiac pathological conditions and examine regulatory interactions with splicing factors using RNA-seq data. Methods: Three human heart RNA-seq databases (PRJNA198165, PRJNA477855, PRJNA678360) were used, comprising various cardiac conditions like non-ischemic cardiomyopathy (NICM), ischemic cardiomyopathy (ICM), dilated cardiomyopathy (DCM), and heart failure with reduced ejection fraction (HFrEF), with and without left ventricular assist device (LVAD) support. Splicing events were identified using the rMATS tool, and correlation analyses were performed between SRF and predicted splicing factors. Functional enrichment of SRF-correlated genes was assessed via Gene Ontology (GO) and KEGG pathways. Results: The skipped exon (SE) events were the predominant splicing type across all datasets. SRF chr6, including (Exon 2, 43,173,847–43,174,113), (Exon 4, 43,176,548–43,176,667), and (Exon 5, 43,178,294-43,178,485), were most frequently involved in SE and mutually exclusive exon (MXE) events across multiple heart failure subtypes. Correlation analysis revealed strong positive associations between SRF and several splicing factors (HNRNPL, HNRNPD, SRSF5, and SRSF8). GO and KEGG analyses revealed enrichment of muscle development, sarcomere structure, lipid metabolism, and immune signaling pathways. Conclusions: Our study shows that SRF is subject to extensive alternative splicing in heart failure, particularly at Exon 2 and Exon 5, suggesting isoform-specific roles in cardiac remodeling. The strong co-expression with specific splicing factors delineates a regulatory axis that may explain the pathological transcriptome in cardiomyopathy. These findings provide a foundation for exploring splicing-based biomarkers and therapeutic targets in cardiac pathology for SRF. Full article

Background: Arrhythmogenic cardiomyopathy (ACM) is an inherited heart disorder characterized by structural and functional myocardial alterations, often accompanied by ventricular arrhythmias (VAs), which may ultimately result in sudden cardiac death (SCD). While mutations in genes coding for desmosomal components are commonly identified in affected individuals, genetic variants involving non-desmosomal proteins have recently been recognized as contributors to the disease’s etiology. In 2008, a mutation in the transmembrane protein 43 (TMEM43) was identified as being responsible for a fully penetrant, sex-related, and severe form of ACM. This review aimed to systematically synthesize the current evidence on the natural history, electrocardiographic, and imaging findings as well as the clinical outcomes of TMEM43 cardiomyopathy. Methods: A systematic search was performed in the PubMed, Scopus, and Web of Science databases, following the PRISMA guidelines, using the terms “TMEM43” AND “cardiomyopathy”. After an initial screening of 144 retrieved articles, 80 were considered relevant. Upon a full-text review and eligibility assessment, 12 studies involving 903 individuals harboring TMEM43 variants were selected for inclusion. Results: Male patients more frequently carried the pathogenic TMEM43 variant (n = 505, 55.9%) and exhibited an earlier arrhythmic onset of the disease (33.2 years old versus 46.2 years old in female patients), supporting the need for earlier implantable cardioverter–defibrillator implantation (30.4 versus 42.2 years old). Palpitations, chest pain, and syncope were the most common presenting symptoms. Baseline electrocardiograms commonly demonstrated poor R wave progression, QRS prolongation, and premature ventricular contractions (PVCs). Arrhythmic events, including malignant VAs and SCD, were early manifestations of the disease, especially in male patients. Frequent PVCs and left ventricular dilation were considered early markers of the disease and were predictive of arrhythmic events. Conversely, heart failure was reported as a late clinical outcome, requiring heart transplantation in a minority of cases (1.5%). Conclusions:TMEM43 cardiomyopathy is a fully penetrant autosomal dominant form of ACM, characterized by a well-defined clinical phenotype that is more severe and presents earlier in male patients. Full article

Background: Leigh’s Disease is a rare mitochondrial disorder primarily affecting the central nervous system, with frequent secondary cardiac manifestations such as hypertrophic and dilated cardiomyopathies. Early detection of cardiac complications is crucial for patient management, but manual interpretation of cardiac MRI is labour-intensive and subject to inter-observer variability. Methodology: We propose an integrated deep learning framework using cardiac MRI to automate the detection of cardiac abnormalities associated with Leigh’s Disease. Four CNN architectures—Inceptionv3, a custom 3-layer CNN, DenseNet169, and EfficientNetB2—were trained on preprocessed MRI data (224 × 224 pixels), including left ventricular segmentation, contrast enhancement, and gamma correction. Morphological features (area, aspect ratio, and extent) were also extracted to aid interpretability. Results: EfficientNetB2 achieved the highest test accuracy (99.2%) and generalization performance, followed by DenseNet169 (98.4%), 3-layer CNN (95.6%), and InceptionV3 (94.2%). Statistical morphological analysis revealed significant differences in cardiac structure between Leigh’s and non-Leigh’s cases, particularly in area (212,097 vs. 2247 pixels) and extent (0.995 vs. 0.183). The framework was validated using ROC (AUC = 1.00), Brier Score (0.000), and cross-validation (mean sensitivity = 1.000, std = 0.000). Feature embedding visualisation using PCA, t-SNE, and UMAP confirmed class separability. Grad-CAM heatmaps localised relevant myocardial regions, supporting model interpretability. Conclusions: Our deep learning-based framework demonstrated high diagnostic accuracy and interpretability in detecting Leigh’s disease-related cardiac complications. Integrating morphological analysis and explainable AI provides a robust and scalable tool for early-stage detection and clinical decision support in rare diseases. Full article