Computer Vision Analytics Engineer – Medical Video/Image Analytics
Job Description:
We are seeking Computer Vision Analytics Engineers to support a Medical Video Analytics Project. This initiative integrates real-time
medical video processing, AI-powered computer vision, and cloud-based analytics to enhance endoscopic procedures and MRI
imaging.
The role involves working on edge-to-cloud video processing pipelines, developing vision algorithms for real-time object detection,
and building machine learning models that generate automated insights and recommendations for medical professionals.
Key Responsibilities:
• Work with real-time video feeds from robotic-assisted surgery and endoscopic procedures.
• Support remote and in-hospital control workflows for AI-enhanced video analytics.
• Process and analyze high-speed medical video streams at gigabit-per-second (Gbps) throughput.
• Ensure secure transmission of MRI and endoscopic video feeds from edge devices to the cloud.
• Develop scalable Edge-to-Cloud AI solutions, ensuring low-latency inference for various medical applications.
• Implement AI models that analyze video content and classify frames as useful or non-useful.
• Develop AI-driven video segmentation and classification models to filter relevant vs. non-relevant frames.
• Develop object detection, segmentation, and tracking models to identify anatomical structures, surgical instruments, and
procedural steps in real time.
• Implement video enhancement and denoising techniques to improve image clarity and feature extraction.
• Deploy deep learning-based models for medical video analytics using TensorFlow, PyTorch, and OpenCV.
• Compare real-time footage with pre-trained medical video datasets to generate automated insights.
• Develop containerized AI models (Docker, Kubernetes) to ensure scalable deployment in hospital environments.
• Integrate AI-powered video analytics pipelines with cloud-based AI models (e.g., Azure AI)
• Ensure seamless bi-directional communication between cloud AI models and edge computing systems.
• Work closely with radiologists and healthcare professionals to fine-tune AI-driven video object detection and
recommendations.
• Integrate AI-powered video analytics solutions with existing hospital PACS, DICOM storage, and medical imaging
infrastructure.
• Ensure AI models comply with HIPAA, FDA, and medical device regulations for clinical deployment.
Requirements
Qualifications:
• Demonstrated experience in computer vision, AI model development, and optimization.
• Experience working with medical videos, including MRI, endoscopy, ultrasound, echo-cardiograms, and OCR-based
recognition.
• Proficiency in multi-modal AI, integrating various medical imaging sources.
• Experience working closely with healthcare professionals and hospital workflows.
• Experience integrating AI models with hospital IT systems, PACS, and DICOM-based workflows.
• Proficiency in Python and experience with AI frameworks such as PyTorch, TensorFlow, OpenCV.
• Expertise in computer vision techniques, including Object detection (YOLO, SSD, Faster R-CNN), Image segmentation (U-Net,
Mask R-CNN), Image classification (ResNet, EfficientNet, ViTs), Feature extraction (SIFT, SURF, ORB)
• Strong knowledge of machine learning techniques including Supervised, unsupervised, and self-supervised learning, CNNs,
Vision Transformers (ViTs), GANs, attention-based networks, Random forests, SVMs, boosting algorithms
• Proficiency in data preprocessing, augmentation, normalization, and handling large-scale image datasets.
• Experience working with multi-GPU workloads for training and inference.
• Experience deploying models using containerization technologies (Docker, Kubernetes).
• Experience with high-performance computing (HPC) techniques for managing large-scale datasets.
• Background in federated learning for medical AI to enhance privacy-preserving model training.
• Prior experience in developing AI solutions for real-time clinical applications.
• Strong understanding of regulatory constraints in AI-driven medical applications.
• Ability to effectively communicate complex AI models to technical and non-technical stakeholders.