Razeem Ahmad

Razeem Ahmad

Data Scientist
India

About Me

Data Scientist with experience in computer vision, deep learning, federated learning, and data science. Worked on cardiac MRI and brain MRI projects, built dashboards and analytics solutions, and has experience with Pyth…

Experience

Data Scientist

Philips, Bangalore
Jul 2022 - Present · 4 years

Led project in cardiac MRI on automating end-to-end patient scan session.
Worked on computer vision algorithms including 3D segmentation, automated landmark localization, and registration.
Used deep learning techniques including HighResNet and nnUNet.
Worked on federated learning for on-site learning.
Built scalable dashboards in Power BI and Hadoop to provide statistical solutions to doctors and clinical experts.
Simplified complex and large amounts of data into business decisions.
Accelerated the project completion process from 10 to 3 months.
Automated the whole annotation pipeline.
Reduced dependency on external contractors.
Reduced effort on clinical expert verification, cutting down time to one-fourth per scan.
Patent pending.

Data Analyst

Philips
Jan 2021 - Present · 5 years 6 months

I am a Data scientist, graduated from India's #1 ranked Technology institute IIT Madras, specialised in Biomedical and Al. I have good communication in English and Arabic.

Have been working on Big Data, AWS and AI in healthcare. Headed the data side of project in Cardiac MRI focused in predicting oblique views of the scanned heart. This required obtaining loads of manual annotations from which terabytes of secondary data had to be processed like annotator variability, data confidence, view significance (with Wald test, t-tests) etc.

Built systems based on machine learning algorithms (PCA, tSNE, genetic algorithm) to optimize parameters. Developed deep learning model (nnUNet, MRNet, SVD based registration) to predict cardiac views(94% score, patent pending), classify quality of scans (F1 score 0.9) and adapt to end user preferences (MSE 3degrees) respectively

Deep Learning Intern

Philips
Dec 2020 - Jun 2022 · 1 year 6 months

Worked on Brain MRI scan data to localize slices employing 3D U-Net framework.
Applied Gradient-weighted Activation Mapping (Grad-CAM) to produce a coarse localization map.
Released publications for the project.
Enhanced the existing automatic planning algorithm using Single Value Decomposition (SVD).
Predicted slices from 3D point cloud from just 20 training data.
Achieved angulation and intercept error of 3.5 deg and 2mm respectively.

Business Development Intern

Elan STEMcademy, Bangalore
May 2018 - Dec 2018 · 7 months

Developed dashboards and revenue projections to suggest profitable decisions.
Pushed to start franchising.
Improved start-up profits by 12%.

Data science intern

Linecraft AI

Implemented code for statistical analyses and visualizations for a manufacturing line in in-house analytics software.
Applied regression models to forecast valuations based on continuously acquired real-time data.
Implemented pipeline to query, mine and clean data.
Performed feature engineering and feature selection.

Skills

Analytics Data Analysis Professional Information Systems Management Data Mining Microsoft Access Statistical Analysis System (SAS) Structured Query Language (SQL) Data Representation Data Analysis Skills Power BI R Programming Language (R) Tableau Linux OS Linux/Ubuntu Windows MySQL Postgres Python OpenCV scikit ANTs PyTorch TensorFlow Spark Git Data Studio AWS Amazon SageMaker Excel Deep Learning Computer Vision Federated Learning 3D Segmentation Automated Landmark Localization Registration 3D U-Net Grad-CAM SVD Regression Models Feature Engineering Feature Selection Statistical Analysis Machine Learning SVM CNN
Report this Profile?