Job Description

About Us:  

At RNT Health Insights, our mission is to revolutionize gastrointestinal diagnostics through our innovative Software as a Medical Device (SaMD) solutions. Our technology can be integrated with existing endoscopic workflows to enhance diagnostic accuracy by identifying, characterizing, and delineating lesions that are often missed or misdiagnosed, in real-time during endoscopic procedures. Our flagship SaMD product focuses on early gastric cancer detection, significantly improving the precision and efficiency of diagnosing early-stage gastric cancer and associated malignant lesions in real-time during endoscopic procedures.

Job Description:

As a Computer Vision Engineer, you will play a pivotal role in researching, developing, and implementing spatiotemporal algorithms for detecting and delineating lesions in real-time during endoscopic video feeds. You will collaborate with a multidisciplinary team of AI researchers, gastroenterologists, and software engineers to enhance both spatial and temporal accuracy in lesion detection. This role involves designing preprocessing pipelines, optimizing models, and ensuring low latency, high-performance deployment in clinical settings.

Key Responsibilities:

•       Algorithm Development: Research, develop, and implement spatiotemporal techniques combined with CNN and other spatial models for real-time lesion detection in endoscopic video streams.

•       Temporal Analysis: Investigate and apply state-of-the-art techniques (e.g., LSTMs, 3D CNNs) to model temporal dependencies in video-based data for improved lesion tracking and detection.

•       Model Integration: Integrate temporal models with existing CNN-based spatial models to create efficient, end-to-end pipelines for real-time inference during endoscopy procedures. • Preprocessing & Inference Pipelines: Design and implement robust video preprocessing (e.g., frame extraction, image enhancement, noise reduction) and inference pipelines that ensure smooth integration with endoscopic hardware and software systems.

•       Post-processing Optimization: Work on post-processing techniques to improve lesion localization, classification, and segmentation accuracy, and ensure consistent performance in different clinical settings.

•       Model Optimization: Fine-tune models for deployment on constrained hardware platforms, ensuring low-latency performance without compromising accuracy.

•       Collaborative Research: Collaborate with research teams to explore and incorporate cutting-edge temporal models, multi-frame fusion techniques, and domain-specific innovations in medical video analysis.

•       Performance Benchmarking: Benchmark models against datasets, assess performance (accuracy, speed, robustness), and optimize for real-time clinical use.

Job Tags

Similar Jobs