Project 01 · Machine Learning

X-Ray Image Classification & Diagnostic Insight Tool

Machine learning pipeline for analyzing X-ray images and highlighting patterns associated with common abnormalities.

Developed a Python-based medical imaging project to analyze and classify X-ray images for common abnormalities, with a focus on fractures and lung-related findings. The project included image preprocessing, feature extraction, model training, and output visualization designed to make predictions more interpretable and easier to review.

Highlights

  • Built an image analysis workflow for classifying radiographic abnormalities from publicly available X-ray datasets.
  • Applied preprocessing and feature extraction methods to improve model input quality and support training performance.
  • Visualized prediction outputs and model behavior to make results more interpretable in a healthcare-focused context.

Workflow components

  • Dataset preview, normalization, and augmentation
  • Feature extraction and classification pipeline
  • Confusion matrix and interpretability outputs
  • Correct and incorrect classification review

Tools

Python Machine Learning Medical Imaging Image Preprocessing Data Visualization
Project 02 · 3D Slicer

3D CT Anatomical Segmentation in 3D Slicer

3D segmentation and reconstruction workflow for converting CT imaging data into interpretable anatomical models.

Built a 3D medical imaging case study in 3D Slicer focused on segmenting CT-based anatomical structures and generating 3D reconstructions for improved visualization. The workflow involved importing DICOM scans, isolating regions of interest, refining segmentation masks, and producing 3D renderings that make anatomical relationships easier to understand for education and imaging review.

Highlights

  • Designed a segmentation workflow in 3D Slicer to transform CT imaging data into clear 3D anatomical reconstructions.
  • Used region-of-interest isolation and mask refinement techniques to improve structure visibility and anatomical clarity.
  • Produced interactive and static visual outputs suitable for imaging education, case presentation, and spatial anatomy review.

Workflow components

  • Original CT slices and DICOM import
  • Segmented structure overlays with thresholding
  • Mask refinement and smoothing
  • Side-by-side 2D and 3D rendered views

Tools

3D Slicer DICOM Segmentation Volume Rendering 3D Modeling
Project 03 · OsiriX

OsiriX DICOM Review and Multi-Planar Imaging Workflow

Structured DICOM review process for navigating studies across axial, sagittal, and coronal views.

Created a structured imaging review workflow in OsiriX for navigating DICOM studies across multiple planes, organizing image series, and identifying clinically relevant visual patterns. The project emphasized efficient study review, image annotation, and consistent inspection across axial, sagittal, and coronal views to strengthen radiologic interpretation and workflow familiarity.

Highlights

  • Built a repeatable OsiriX-based workflow for reviewing DICOM studies across multiple anatomical planes.
  • Organized imaging series and annotated regions of interest to support structured visual analysis.
  • Strengthened familiarity with radiology-style image review workflows and cross-sectional anatomical interpretation.

Workflow steps

  • Import and organize DICOM series
  • Establish multi-planar layout across axial, sagittal, coronal
  • Annotate regions of interest with consistent conventions
  • Review intensity profiles and cross-sectional alignment

Tools

OsiriX DICOM Multi-Planar Reconstruction Annotation Imaging Review
Project 04 · 3D Slicer

Lesion Segmentation and Volumetric Analysis in 3D Slicer

Case-study workflow for segmenting lesions and generating reproducible 2D and 3D measurement outputs.

Developed a lesion-focused imaging case study in 3D Slicer involving segmentation of a selected region of interest, volumetric measurement, and comparison between slice-based and 3D views. The project was designed to explore reproducible measurement practices and demonstrate how structured visualization can support more consistent review of lesion shape, location, and size.

Highlights

  • Segmented a lesion-focused region of interest in 3D Slicer to study structure boundaries and spatial morphology.
  • Generated volumetric and dimensional measurements to compare lesion appearance across 2D slices and 3D reconstructions.
  • Built a case-study workflow centered on clear visualization, reproducibility, and structured image-based analysis.

Workflow components

  • Original lesion view and slice selection
  • Segmented ROI with refined mask
  • Volumetric and dimensional measurements
  • Reproducibility notes and clinical-context commentary

Tools

3D Slicer Segmentation ROI Analysis Volume Measurement 3D Visualization

Looking for more detail?

Reach out to discuss specific case studies, methodology, or imaging workflow design in more depth.