• Versatile Senior Data Scientist with a Ph.D. (pursuing) in Image Processing and over 7 years of experience in developing machine learning and deep learning models across various industries.
• Expert in creating data-driven solutions for complex problems, with a strong background in computer vision, artificial intelligence, and advanced analytics.
• Experienced in using Python, TensorFlow, and other tools to build scalable models that drive innovation and efficiency.
• Passionate about applying cutting-edge technology to solve real-world challenges and improve business outcomes.

Programming Skill: Python, Matlab
Deep Learning and Machine Learning Toolkits: Tensorflow, Keras, Sklearn, MATLAB Deep Learning Toolbox
Data Visualization: Pandas, Seaborn

Timeline: June 2021 - Present
Role: Senior Data Scientist, Bengaluru

Ischemic Stroke Segmentation

• Developed a CNN-based model for the segmentation of Ischemic Stroke in multiparametric MRI, addressing challenges posed by the varying sizes of affected brain regions and temporal changes in their appearance.
• Enhanced model performance by incorporating physics-based features into the training data.
• Deployed the model on an inference engine for processing retrospective scans.

Denoising of Low-Field MRI Scans

• In MRI, higher magnetic fields provide better signal-to-noise ratios (SNR) in images, but the equipment is expensive. To make high-quality MRI more accessible, low-field MRI systems with lower specifications are being introduced, where reduced SNR is compensated by AI-based reconstruction.
• Developed CNN-based algorithms to denoise images from low-field MRI, enhancing them to diagnostic quality. This involved training densely connected architectures with naturally simulated images and testing them on MRI images obtained with lower specifications.

Cloud-Based Tool to Monitor the Research Scanner

• For a research scanner, it's essential to monitor every scan, as the scan data is crucial for transitioning research into a product.
• Developed a web-based tool for remote monitoring to ensure adherence to study protocols and regulatory guidelines.

Timeline: June 2020 - June 2021
Role: Technical Lead, Chennai

X-ray Analysis Tool for Monitoring Respiratory Disease

• Directed the design, development, and implementation of deep learning algorithms for analyzing chest X-rays. The models were trained to accurately detect and monitor various respiratory diseases, enhancing diagnostic capabilities in clinical settings.
• Managed a multidisciplinary team of data scientists and software engineers, ensuring timely delivery and adherence to high standards for medical applications.
• Provided technical guidance and mentorship, fostering a collaborative environment that encouraged innovation and problem-solving.

Ph.D. Thesis
Towards Developing Deep Learning Algorithms for Brain Lesion Segmentation

• The thesis focuses on developing deep learning algorithms for the challenging task of brain lesion segmentation, specifically ischemic stroke segmentation and brain tumor segmentation.
• Automated segmentation of ischemic lesions from CT perfusion maps using an encoder-decoder fully convolutional neural network, following pre-processing steps like skull stripping and perfusion map standardization.
• Utilized 3D fully convolutional neural networks with dense connectivity and residual connections for glioma segmentation from multimodal MRI, employing hard mining to improve segmentation by increasing the Dice Similarity Coefficient (DSC) threshold for challenging cases.

Segmentation of Kidney and Kidney Tumor from CT Scan of the Kidney (2019):

• Developed a CNN-based segmentation network for fully automatic segmentation of kidneys and tumors from CT images, as part of the KiTS2019 Challenge at MICCAI.

3D CNNs for Molecular Subtype Prediction in Glioblastoma Multiforme (2019):

• Showcased deep learning models to identify associations between brain imaging phenotypes and molecular subtypes, using MRI images of Glioblastoma Multiforme from The Cancer Genome Atlas for molecular subtype classification.

Detection of Pneumonia on Chest X-ray (2018):

• CNN-based algorithms have been built to detect a visual signal for pneumonia by automatic localization of lung opacities in chest radiographs.
• This project was part of the RSNA Pneumonia Detection Challenge held on Kaggle. Our algorithms secured a position in the top 8%.

Identification of Myelomatous Lesions from Multi-Parametric MR Images of Pelvic Bones (2018):

• Several machine learning models such as MLP, SVM, Random Forest, etc., were developed to determine the type of myeloma using features such as first-order and second-order (GLCM) extracted from radiological scans.

• Selvam, Minmini, Anupama Chandrasekharan, Abjasree Sadanandan, Vikas K. Anand, Sidharth Ramesh, Arunan Murali, and Ganapathy Krishnamurthi. "Radiomics analysis for distinctive identification of COVID-19 pulmonary nodules from other benign and malignant counterparts," Scientific Reports 14, no. 1 (2024): 7079.
• Selvam, Minmini, Anupama Chandrasekharan, Abjasree Sadanandan, Vikas Kumar Anand, Arunan Murali, and Ganapathy Krishnamurthi. "Radiomics as a non-invasive adjunct to Chest CT in distinguishing benign and malignant lung nodules," Scientific Reports 13, no. 1 (2023): 19062.
• Anand, Vikas Kumar, et al. "Brain tumor segmentation and survival prediction using automatic hard mining in 3D CNN architecture." Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II 6. Springer International Publishing, 2021.
• Acharya, Gagan, et al. “3D Convolution Neural Networks for Molecular Subtype Prediction in Glioblastoma Multiforme.” Medical Imaging 2019: Imaging Informatics for Healthcare, Research, and Applications, 2019.
• Ignatov, Andrey, et al. “NTIRE 2019 challenge on image enhancement: Methods and results.” In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, pp. 0-0, 2019.
• Anand, Vikas Kumar, et al. “Fully Automatic Segmentation for Ischemic Stroke Using CT Perfusion Maps.” Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries Lecture Notes in Computer Science, 2019, pp. 328–334.
• Anand, Vikas Kumar, et al. “Semi-automatic identification of myelomatous lesions from multi-modal MR images.” European Congress of Radiology 2018.

• Teaching Assistant for the NPTEL Course on Machine Learning for Engineering and Science Applications (Jan-Oct 2019)
• Teaching Assistantship for Courses in the Department of Engineering Design at Indian Institute of Technology MadrasTeaching
• Assistant for Digital Signal Processing (ED5017) (Jan-May 2020, Jan-May 2019, Jan-May 2018)
• Teaching Assistant for Medical Image Analysis (ED6001) (July-Nov 2019, July-Nov 2018)
• Teaching Assistant for Machine Learning for Engineering and Science Applications (ID5030) (Jan-May 2018)
• Literary Secretary, Bhadra Hostel, IIT Madras (2017–2018)

• (CS5011) Introduction to Machine Learning
• (CS7015) Deep Learning
• (EE5175) Image Signal Processing
• (EE5121) Optimization Methods in Signal Processing and Communication
• (AM5160) Biomedical Imaging System
• (CY6101) Magnetic Resonance Imaging