Ongoing Research projects in MAL Lab
Motion and Force Planning for Manipulating Heavy Objects by Rolling
A common way to manipulate heavy objects is to maintain at least one point of the object in contact with the environment during the manipulation. When the object has a cylindrical shape or, in general, a curved edge, not only sliding and pivoting motions but also rolling the object along the edge can effectively satisfy this condition. Edge-rolling offers several advantages in terms of efficiency and maneuverability. This paper aims to develop a novel approach for approximating the prehensile edge-rolling motion on any path by a sequence of constant screw displacements, leveraging the principles of screw theory. Based on this approach, we proposed an algorithmic method for task-space-based path generation of object manipulation between two given configurations using a sequence of rolling and pivoting motions. The method is based on an optimization algorithm that takes into account the joint limitations of the robot. To validate our approach, we conducted experiments to manipulate a cylinder along linear and curved paths using the Franka Emika Panda manipulator.
Publication:
- M. Boroji, V. Danesh, I. Kao, A. Fakhari, “Motion Planning for Object Manipulation by Edge-Rolling”, 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024 (Accepted for Publication).
Ex-Vivo Raman Spectroscopy of Soft Tissue Sarcoma
Soft tissue sarcomas (STS) are a diverse and rare group of malignant tumors arising from the connective tissues of the body, including muscles, fat, nerves, and blood vessels. The heterogeneity and infrequency of these tumors pose significant challenges in both diagnosis and treatment. Surgical resection remains the primary treatment strategy, often complemented by radiation or chemotherapy, contingent upon the tumor’s size, location, and stage. However, current methods for assessing intraoperative margins are limited, underscoring the need for improved approaches that enhance both efficiency and accuracy. This study investigates the potential of microscopic Raman spectroscopy for distinguishing between STS, benign tumors, and normal tissue types. Ex-vivo Raman measurements were conducted using a 633 nm excitation wavelength on samples obtained from surgical resections of 28 patients (more than 700,000 spectra). After pre-processing of the data and outlier detection, a classification algorithm was developed to accurately categorize the different tissue types, achieving an overall weighted accuracy of 92%. These findings suggest that single Raman spectra could serve as a rapid, non-invasive tool for surgical guidance, aiding in the precise identification of abnormal tissue margins and types.
In Vivo Skin Cancer diagnosis using Raman Spectroscopy
- Real time Raman Spectroscopy to diagnose skin cancer and margins using Neural Networks Analysis
- Convolutional Neural Network Attention-Guided Segmentation to Identify the Edges of Benign and Malignant Skin Tumors
Publication:
- M. Boroji, V. Danesh, P. Prasanna, J. Kim, X. Qian, M. Khari, K. Mani, B. F Boyce, F. A Khan, S. Ryu, I. Kao, R. F Cattell, “Convolutional Neural Network Attention-Guided Segmentation to More Accurately Identify the Edges of Benign and Malignant Skin Tumors”, 2024 AAPM 66th Annual Meeting & Exhibition
Fracture Reduction of Femur bone using Computer Vision and Image Guided Techniques