Freelance data scientist · All case studies

Four MRI stories: glioma, meningioma, pituitary, none.

A computer-vision pipeline on brain MRI: four clinical classes, images loaded and resized with OpenCV and PIL into standardised arrays, ready for a deep net that has to tell tumour type from a healthy scan.

Client: Neuroimaging study. Built by Dilshad Raza.

Glioma, meningioma, and pituitary masses do not look like the same blob, and a healthy scan has to stay in the negative class. The images arrived at mixed sizes. Before any architecture could train, the set had to be a tensor the net would not reject.

I sourced a comprehensive brain-tumour MRI set, mapped the four clinical labels, and built the loading pipeline in OpenCV and PIL — resize, array, stack — so deep architectures saw one shape. Classification sat on that standardised tensor, not on a folder of mismatched JPEGs.

Hire Dilshad Raza for similar freelance data science, machine learning, and AI automation in the UK, United States, and Australia.