MekaNet: A Deep Learning Framework for Megakaryocyte Detection and Myeloproliferative Neoplasm Classification with Enhanced Feature Engineering
A submitted MekaNet manuscript connecting megakaryocyte detection, MPN classification, and clinically shaped feature engineering.
A medical-image manuscript on megakaryocyte detection and myeloproliferative neoplasm classification with enhanced feature-engineering support.
problem
Megakaryocyte detection and MPN classification require clinically meaningful image features, reliable detection, and evaluation beyond generic computer-vision metrics.
key idea
Combine detection-oriented deep learning with feature-engineering choices that better reflect the pathology task.
my role
Technical contributor for model/evaluation workflow and manuscript-support artifacts.
methods
- • Deep learning detection
- • Feature engineering
- • Pathology image analysis
- • Clinical classification workflow
evidence / results
- • Submitted manuscript identified in the Overleaf paper-candidate index
- • Part of the broader MekaNet medical AI line
why this belongs in the portfolio
- • Strengthens the applied-medical-AI axis
- • Shows how domain constraints reshape model and evaluation design
authors
Jae-Hyun Baek et al.
venue / status
Submitted medical AI manuscript
Submitted/draft manuscript; public detail intentionally remains high-level until venue status is final.
tags