Memuat
Memuat
Build system pneumonia detection from chest X-ray with PyTorch, from dataset to clinical evaluation dan Grad-CAM. This course follows a real research workflow, baseline model, proposed model hybrid CNN-LSTM, sensitivity-specificity–ROC-AUC metrics, so it can directly become a thesis, dissertation, or publication framework.
Swipe to view curriculum, architecture, and model results.
Sensitivity of CNN-LSTM on Test Data
ROC-AUC (CNN baseline 0,966)
Best model validation accuracy
Training time on CPU, no GPU
Each stage of medical image AI research is practiced directly with PyTorch code.
PneumoniaMNIST (MedMNIST): 5.856 chest X-rays of children with standard train/validation/test split for research.
CLAHE for local contrast, normalization, rotation-zoom-shift-brightness augmentation with OpenCV, Dataset & DataLoader PyTorch.
Conv2d, BatchNorm, dropout, training loop, and required comparison model in every research.
CNN feature maps are read as a sequence by bidirectional LSTM, from top to bottom of the lungs.
Class weights for imbalanced data, ReduceLROnPlateau, early stopping, and best model checkpoint.
Sensitivity, specificity, ROC-AUC, Youden threshold, and Grad-CAM to see the area noticed by the model.
9–15 minute videos per module, complete with PDF guides. Module 1 can be watched for free.
Methodology framework, code, and results that can directly form the basis of chapters 3 and 4.
Complete and reproducible pipeline for medical image experiments, easily replaceable dataset or architecture.
Understand how AI diagnosis works, clinical metrics, and its limitations.
Intermediate-level PyTorch practice: hybrid architecture, Grad-CAM hooks, and model evaluation.
We help with model implementation, experiments, evaluation, and result writing for theses, dissertations, or publications.
Lihat Jasa Riset CNN-LSTM Citra MedisA combination of Convolutional Neural Network that extracts spatial features from images, and Long Short-Term Memory that reads these features as a sequence to capture relationships between different parts of the image.
No. 64x64 pixel images make the entire training complete under one minute on a regular laptop CPU.
PneumoniaMNIST from the MedMNIST collection (CC BY 4.0 license), a derivative of the chest X-ray dataset by Kermany et al. This dataset is widely used in publications and can be used for research with proper citation.
Yes. Premium Class customers and class buyers receive a zip file containing all source code, dataset, preprocessed data, and trained model.
Yes. The same pipeline can be used for other medical images such as retina, skin, or CT by changing the dataset and adjusting the number of classes.
Yes, live online classes and also AI medical image research support services are available for thesis, dissertation, or publication.
Consult your research topic, we help you choose the most suitable learning package.
Free Consultation