A method for classifying medical images using transfer learning: A pilot study on histopathology of breast cancer
importance 6/10
ai assisted
AI in cancer detection oncology imaging stage: pilot
Problem
Detecting breast cancer automatically from histopathology images.
Why it matters
Breast cancer detection accuracy directly affects patient diagnosis and outcomes.
What the AI did
Transfer learning fine-tunes Google's Inception v3 model, pre-trained on non-medical images, to classify breast cancer histopathology images.
Results
AUC of 0.93
Impact
Pilot study demonstrates feasibility of transfer learning for breast cancer histopathology classification.
Record
| Venue | IEEE International Conference on e-Health Networking, Applications and Services (Healthcom) |
|---|---|
| Published | 2017 |
| Work type | conference-paper |
| Authors | Jongwon Chang, Jisang Yu, Taehwa Han, Hyuk‐Jae Chang, Eunjeong Park |
| Citations | 115 (OpenAlex, at import time) |
| References | 20 |
| Model or method | Google Inception v3 (transfer learning) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1109/healthcom.2017.8210843 |
| Open access | closed |
| Keywords | breast cancer, histopathology, transfer learning, Inception v3, medical imaging |
Importance rating
6
significant — thousands to millions affected
Breast cancer detection accuracy directly affects patient diagnosis and outcomes, a major disease area, but the abstract is a pilot feasibility study with no scale stated.