Showing papers by "Debdoot Sheet published in 2022"

I’m GROOT: a multi head multi GRaph netwOrk recognizing surgical actiOn Triplets✱

Abstract: Laparoscopic cholecystectomy is a widely performed minimally invasive surgical procedure that imposes many challenges to the operating surgeon. While we strive to understand and automate such surgeries, the key is to identify the actions involved in it. An action involves a set of tools and a target anatomy, together forming the action triplets. However, the relations between the triplets and their constituents are sparse, making it challenging to learn their relations. In this paper, we propose a graph neural network based approach to exploit these underlying sparse relations in the data. We portray the proposed method’s ability to uniformly learn multiple tasks and classify triplets with an mAP of 0.261. In addition, we experimentally show the inability of fully connected and convolution layers to learn these sparse relations when trained on 40 laparoscopic videos and validated using five videos. Codes will be available at : https://github.com/iitkliv/groot.

Verifiable and Energy Efficient Medical Image Analysis with Quantised Self-attentive Deep Neural Networks

Abstract: Convolutional Neural Networks have played a significant role in various medical imaging tasks like classification and segmentation. They provide state-of-the-art performance compared to classical image processing algorithms. However, the major downside of these methods is the high computational complexity, reliance on high-performance hardware like GPUs and the inherent black-box nature of the model. In this paper, we propose quantised stand-alone self-attention based models as an alternative to traditional CNNs. In the proposed class of networks, convolutional layers are replaced with stand-alone self-attention layers, and the network parameters are quantised after training. We experimentally validate the performance of our method on classification and segmentation tasks. We observe a $50-80\%$ reduction in model size, $60-80\%$ lesser number of parameters, $40-85\%$ fewer FLOPs and $65-80\%$ more energy efficiency during inference on CPUs. The code will be available at \href {https://github.com/Rakshith2597/Quantised-Self-Attentive-Deep-Neural-Network}{https://github.com/Rakshith2597/Quantised-Self-Attentive-Deep-Neural-Network}.

Design of a System and Method for Optimal selection of Tumor Slice using Linear Ultrasound Imaging for Histopathology

Abstract: In excision biopsy, a tumor mass is surgically removed from the body. Subsequently, it is sliced at an appropriate location and investigated microscopically through a process called histopathology. Any bias in tumor slicing severely influences histopathology outcomes, such as if malignant foci do not appear in the sliced location, then, the tumor would be accidentally reported non-malignant. The standard approach adopted to solve this challenge by a histopathologist is to overcome this bias by slicing at multiple locations for their investigation. Till now, this process has been manual, time-consuming, and error-prone. We aim to design a system and develop a data-driven deep learning approach to assist histopathologists by providing them with a representative slice location for reporting to increase their efficiency and accuracy. We have developed a low cost linear gantry scanner that can acquire images and integrated with a deep learning model to predict the optimal slice representative of pathology in a tumor mass. We achieve an F1 score of 0.97 and an accuracy of 97.5% in predicting an optimal slice using this approach.