20 Nov
|
Arohana Tech Solutions Private
|
New Delhi
20 Nov
Arohana Tech Solutions Private
New Delhi
Key Responsibilities
– Deep Learning Research Engineer – Medical Imaging
- Develop deep learning models for 2D/3D image segmentation, including tumor, lesion, and organ segmentation.
- Preprocess medical images, including DICOM/NIfTI handling, resizing/resampling, normalization, cropping, and augmentation.
- Explore, implement, and benchmark deep learning and foundation models for oncology-related medical image segmentation, classification, and lesion detection.
- Perform transfer learning, fine-tuning, feature extraction, and controlled ablation experiments.
- Train, fine-tune, and evaluate models using Python, PyTorch/TensorFlow, MONAI, and nnU-Net.
- Evaluate models using appropriate metrics, including Dice/IoU for segmentation and accuracy, precision, recall, F1-score, ROC-AUC, and calibration for classification.
Experience
- 4–7 years of hands-on experience in developing and fine-tuning AI/ML models.
- Practical exposure to NLP, computer vision, or predictive analytics projects.
- Demonstrated experience in implementing AI solutions using open-source or cloud frameworks
Technical Competencies
Understanding Of Foundational Deep Learning Models And Architectures, Including
- Image segmentation: U-Net, U-Net++, Attention U-Net, 3D U-Net, nnU-Net, and DeepLab.
- Vision Transformers: ViT, DeiT, Swin Transformer, and Swin-UNETR.
- Foundation and self-supervised models: SAM/MedSAM and DINOv2 for transfer learning or feature extraction.
- CNN–Transformer hybrid architectures: TransUNet, UNETR, Swin-UNETR, and hybrid CNN/Transformer encoders.
- Detection and instance segmentation: Mask R-CNN, especially for lesion detection or digital pathology.
- Classification: CNN-based classifiers, ResNet/EfficientNet-style backbones, ViT/Swin-based classifiers, and transfer-learning approaches.
Educational Qualification
Bachelor’s/Master’s degree in Computer Science, AI/ML, Biomedical Engineering, Electrical Engineering, Medical Imaging, or a related field.
- - Solid Python programming skills.
- Hands-on experience with PyTorch or TensorFlow; PyTorch is preferred.
- Understanding of CNNs, transfer learning, data augmentation, and standard training/evaluation workflows.
- Foundational deep learning concepts: Strong understanding of neural network fundamentals, including weights and biases, activation functions, forward propagation, loss functions, backpropagation, gradient-based optimization, weight updates, and model training and evaluation.
Skills: cnn,lesion detection,medical imaging,nifti,swin-unetr,image segmentation,swin transformer,deep learning,dicom,monai,pytorch / tensorflow,u-net / nnu-net
📌 Deep Learning Research Engineer – Medical Imaging (New Delhi)
🏢 Arohana Tech Solutions Private
📍 New Delhi