Efficient Deep Learning for Chagas Cardiomyopathydetection: Data-Driven Ecg Feature Reduction

Q: What is the clinical evidence from this study?

Study design: Observational. Population: Chagas cardiomyopathy (n=1630). Intervention: Principal Component Analysis combined with a Convolutional Neural Network (PCA+CNN) vs. Baseline Convolutional Neural Network (CNN) without feature reduction. Primary outcome: Area under the receiver operating characteristic curve (AUC) for classifying Chagas cardiomyopathy.

Q: What does this research mean for the field?

Combining PCA-based dimensionality reduction with a convolutional neural network accurately detects Chagas cardiomyopathy from ECG signals while outperforming baseline deep learning models trained on unreduced data. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Ahmed Faeq Hussein; Auns Qusai Al-Neami; Qais Ahmed Habash

doi:10.46649/ajiea.2025.01103

October 2, 2026Intelligent X-Ray Image Analysis using Fuzzy Inference SystemsOpen Access

Efficient Deep Learning for Chagas Cardiomyopathydetection: Data-Driven Ecg Feature Reduction

View Full Paper

Ask AI

Bookmark

Share

Key result

PCA-CNN pipeline detects Chagas cardiomyopathy with ~97% AUC, outperforming baseline CNN.

n=1,630

Why the study?

Early and efficient diagnosis of Chagas cardiomyopathy using ECGs is limited by the high dimensionality of ECG data, restricting deep learning scalability.

Does a PCA+CNN deep learning model improve the detection of Chagas cardiomyopathy from ECG signals compared to a baseline CNN?

Key Points

To develop a computationally efficient, interpretable deep learning pipeline combining Principal Component Analysis and a Convolutional Neural Network for early Chagas cardiomyopathy detection on ECGs.
Utilized ECG data from the publicly available SaMi-Trop cohort.
Preprocessed signals using a Butterworth high-pass filter to remove baseline wander, followed by PCA for feature space reduction while preserving signal variance.
Trained a custom CNN on the reduced feature space to classify ECG signals as Chagas or non-Chagas, comparing results against an unreduced baseline CNN.
The PCA+CNN model achieved 95.52% sensitivity, 91.11% specificity, 93.31% accuracy, and an AUC of 96.78%, outperforming the baseline CNN (91.40% sensitivity, 89.26% specificity, 91.82% accuracy, and 93.05% AUC).
The PCA+CNN framework yielded a precision of 91.43% and an F1-score of 93.43%, compared to 89.72% precision and 92.00% F1-score for the baseline CNN.
On validation data, the PCA+CNN pipeline correctly identified 246 Chagas and 256 normal cases with 12 false negatives and 24 false positives, compared to 15 false negatives and 29 false positives in the CNN-only model.

Study Design

Type

Observational (n=1,630)

Multicenter

Yes

Structured PICO

Does a PCA+CNN deep learning model improve the detection of Chagas cardiomyopathy from ECG signals compared to a baseline CNN?

P

Population

1,630 ECG recordings (285 normal and 1,345 abnormal) from patients in the Brazilian SaMi-Trop cohort, analyzed for signs of Chagas cardiomyopathy.

I

Intervention

PCA+CNN model (Principal Component Analysis for feature reduction combined with a Convolutional Neural Network for classification)

C

Comparator

Baseline Convolutional Neural Network (CNN) trained on unreduced data

O

Outcome

Diagnostic classification performance (AUC, accuracy, sensitivity, specificity, precision, and F1-score)surrogate

Main Result

Absolute Event Rate: 96.78% vs 93.05%

A deep learning pipeline combining Principal Component Analysis with a Convolutional Neural Network achieved high accuracy and sensitivity for detecting Chagas cardiomyopathy from ECGs while remaining computationally efficient.

Limitations

Real-time application of PCA requires careful handling to ensure consistent transformation between training and inference data
PCA is a linear transformation and may not fully capture complex nonlinear relationships in ECG morphology
PCA introduces an additional preprocessing step that requires consistent transformation during training and deployment

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

May support efficient Chagas cardiomyopathy screening on ECG; leaves open prospective validation before clinical adoption.

Abstract

Chagas cardiomyopathy, a severe complication of chronic Trypanosoma cruzi infection,requires early and efficient diagnosis to improve outcomes. While electrocardiograms (ECGs) areaccessible diagnostic tools, their high dimensionality limits the scalability of deep learning methods. Thisstudy proposes an interpretable and computationally efficient deep learning pipeline combining PrincipalComponent Analysis (PCA) for feature reduction and a Convolutional Neural Network (CNN) forclassification of ECG signals.Using data from the publicly available SaMi-Trop cohort, the ECG signals were preprocessed witha Butterworth high-pass filter to remove baseline wander, followed by PCA to reduce feature space whilepreserving critical variance. A custom CNN was then trained on the reduced feature set to classify signalsas Chagas or non-Chagas.The PCA+CNN model achieved 95.52% sensitivity, 91.11% specificity, 93.31% accuracy, and anAUC of 96.78%, outperforming the baseline CNN trained on unreduced data. The approach reducedoverfitting, improved generalization, and accelerated convergence, demonstrating strong potential for usein real-time and low-resource healthcare settings.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Hussein et al. (2026) conducted an observational in Chagas cardiomyopathy (n=1,630). Principal Component Analysis combined with a Convolutional Neural Network (PCA+CNN) vs. Baseline Convolutional Neural Network (CNN) without feature reduction was evaluated on Area under the receiver operating characteristic curve (AUC) for classifying Chagas cardiomyopathy. The PCA+CNN pipeline achieved an AUC of 96.78%, a sensitivity of 95.52%, and a specificity of 91.11% in detecting Chagas cardiomyopathy, outperforming the baseline CNN model.

synapsesocial.com/papers/6ac72a1436c8cd5d094bf13b https://doi.org/10.46649/ajiea.2025.01103

View Full Paper

Ask AI

Bookmark

Share

Efficient Deep Learning for Chagas Cardiomyopathydetection: Data-Driven Ecg Feature Reduction

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Efficient Deep Learning for Chagas Cardiomyopathydetection: Data-Driven Ecg Feature Reduction

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading