Sujitkumar Gavali

• Validated various algorithms (MATLAB, C++, FPGA) in the MR simulator software application. • Implemented a MR simulator software application in a Linux environment using Windows Subsystem for Linux (WSL), enabling efficient cross‑ platform development and testing.

• Team Neuro‑Inspired AI Inference Acceleration: Proposed a novel magneto‑electric material and a structured process for neuromorphic chip fabrication, designed to significantly enhance the efficiency and scalability of next generation AI models in a sustainable framework.

• Project: Machine learning based detection and quantification of MR artifacts utilizing physics based data simulators & MR Scanners • Generated MRI datasets with and without artifacts using a physics‑based MATLAB MR simulator, enabling the training and validation of deep learning convolutional neural network (CNN) models. • Developed CNN‑based deep learning models for detection and quantification of MR artifacts (type and intensity), enhancing accuracy and efficiency in image processing workflows.

• Project: Visual data generation for multi‑task deep neural networks. • Generated complex datasets from Fashion MNIST and trained convolutional neural networks (CNNs) for multi‑task classification of target im‑ ages.

• Project: Comparison of different B1+ mapping techniques in a phantom and in the human brain using Siemens‑Magnetom 7 Tesla MRI. • Performed 2D–3D image reconstruction using the Actual Flip Angle Imaging (AFI) method to assess radiofrequency (RF) field distribution. • Applied deep learning methods to localizer images to predict 2D relative B1⁺ maps in sub‑seconds.

• Project: Data analysis for the machine learning based prediction of intracranial pressure in epileptic patients. • Analyzed time‑series medical seizure recordings using Python and ML libraries to extract insights and support predictive modeling.

• Debugged and rectified system errors based on logs, user feedback and test tools in Linux environment. • Analyzed and resolved beam physics problems in proton therapy systems as per customer feedback and system generated logs.

• Troubleshot and rectified image quality issues, including MR physics‑related artifacts, image processing, and reconstruction, enhancing diag‑ nostic efficiency and patient throughput. • Resolved critical system errors using logs and DICOM data, created the SOP protocols, remotely and proactively monitored the MR system. • Delivered application training to biomedical engineers, radiographers and radiologists for the safe usage of MRI scanners and fMRI systems.

• Resolved critical LINAC system errors related to X‑ray physics, detectors, and dose performance, improving system efficiency and ensuring smooth patient treatment. • Trained radiation therapists and dosimetrists on safe operation of LINAC, cone beam computed tomoprahy (CBCT), and flat panel detectors in compliance with clinical safety protocols. • Supported doctors, neurosurgeons, physicists with the stereotactic surgical navigation systems which helped to efficiently manage the surgery time.

Dual Degree(HU+TUB) Machine Intelligence - MI Models of Neural Systems - MNS Acquisition and Analysis of Neural Data - AAND Models of Higher Brain Functions - MHBF

Dual Degree(HU+TUB) Machine Intelligence - MI Models of Neural Systems - MNS Acquisition and Analysis of Neural Data - AAND Models of Higher Brain Functions - MHBF

Biomedical Engineering