How flango2023/pulso_f2 works
Pulso Fase 2 is a notebook-based cardiology diagnosis prototype with three documented capabilities: rule-based symptom extraction and disease association, text-based risk classification with subgroup bias evaluation, and binary ECG classification. The README documents these workflows, but no notebook source was provided, so the graph shows only documented relationships.
Main entry points: fase2_parte1_extracao.ipynb Symptom Extraction · sintomas_pacientes.txt Simulated Reports · fase2_parte2_classificador.ipynb TF-IDF Vectors · fase2_iralém2_mlp_ecg.ipynb Binary Class Mapping
Components 15
Symptom Triage
- Simulated Reports
data/sintomas_pacientes.txt - Knowledge Map
data/mapa_conhecimento.csv - Symptom Extraction
notebooks/fase2_parte1_extracao.ipynb - Disease Association
notebooks/fase2_parte1_extracao.ipynb - Extraction Bias Analysis
notebooks/fase2_parte1_extracao.ipynb
Risk Classification
- Labeled Risk Phrases
data/frases_risco.csv - TF-IDF Vectors
notebooks/fase2_parte2_classificador.ipynb - Logistic Regression
notebooks/fase2_parte2_classificador.ipynb - Model Evaluation
notebooks/fase2_parte2_classificador.ipynb - Gender Bias Analysis
notebooks/fase2_parte2_classificador.ipynb
ECG Classification
- MIT-BIH Dataset
- Binary Class Mapping
notebooks/fase2_iralém2_mlp_ecg.ipynb - ECG MLP
notebooks/fase2_iralém2_mlp_ecg.ipynb - Normal or Abnormal
notebooks/fase2_iralém2_mlp_ecg.ipynb
Other components
- Patient Reports
Connections 12
- Patient Reports → Simulated Reports: supplies reports
README.md - Simulated Reports → Symptom Extraction: reads reports
README.md - Knowledge Map → Symptom Extraction: matches symptoms
README.md - Symptom Extraction → Disease Association: associates diseases
README.md - Symptom Extraction → Extraction Bias Analysis: analyzes bias
README.md - Labeled Risk Phrases → TF-IDF Vectors: vectorizes phrases
README.md - TF-IDF Vectors → Logistic Regression: trains model
README.md - Logistic Regression → Model Evaluation: evaluates model
README.md - Model Evaluation → Gender Bias Analysis: audits subgroups
README.md - MIT-BIH Dataset → Binary Class Mapping: provides beats
README.md - Binary Class Mapping → ECG MLP: feeds mapped classes
README.md - ECG MLP → Normal or Abnormal: classifies beats
README.md