Artificial Intelligence for Early Detection and Diagnosis of Cervical Cancer
summary
Claims that CNNs and other deep learning models can flag precancerous and cancerous cervical cells from Pap-smear and whole‑slide images, plus HPV sequence classifiers and mobile digital colposcopy pipelines, backed by experiments using random forest and CNN pipelines (reported 85.1% train / 81.8% test accuracy and a separate case study claiming 95.1% accuracy). The work actually builds image‑and‑genomics pipelines, whole‑slide scanning and mobile screening apps; problems include regional data gaps, need for 5k–10k samples, cost of colposcopes, and demands for explainability and post‑deployment monitoring.
AI-written summary · about
chapters (8)
- 0:00Feminist AI and health justice
- 4:59AI for cervical cancer detection
- 20:43Legal challenges in AI healthcare
- 31:46AI techniques for screening
- 43:03AI limitations in cancer care
- 51:01Funding gaps in women’s health
- 57:23Ethical AI for women’s health
- 1:00:16Closing remarks
related talks
29mCancer diagnostics enabled by machines: finding and classifying single tumor....
55mDeep Learning in Medical Image Diagnostics
34mEnabling Early Stage Breast Cancer Detection Using AI
18mCan AI make us smarter than cancer
24mBreast cancer detection with IBM PowerAI Vision
28mHow AI Can Help Detect Breast Cancer
37mAI in medicine: where do we stand?
21mFighting Cancer with AI