Preliminary exploration on using entropy-weighted hybrid pooling in CNN for ultrasound breast cancer detection - Takeaways - MDSpire

Preliminary exploration on using entropy-weighted hybrid pooling in CNN for ultrasound breast cancer detection

  • By

  • Ratapong Onjun

  • Papon Tantiwanichanon

  • Songkiat Lowmunkhong

  • Tanakorn Sritarapipat

  • Sayan Kaennakham

  • Niwatchai Namwichaisirikul

  • Kitirat Phattaramarut

  • July 8, 2026

  • 0 min

Share

  • 1

    The study introduces an adaptive entropy-weighted hybrid pooling method for CNNs to improve ultrasound breast cancer detection.

  • 2

    Hybrid pooling dynamically combines Max and Average pooling based on local image complexity measured by Shannon entropy.

  • 3

    In the 3-block CNN, hybrid pooling achieved an accuracy of 93.98%, surpassing max pooling's accuracy of 92.72%.

  • 4

    The 4-block CNN showed competitive results with hybrid pooling at 92.90% accuracy compared to max pooling at 94.79%.

  • 5

    The study emphasizes the need for further validation and clinical integration of the proposed adaptive hybrid pooling method.

Original Source(s)

Preliminary exploration on using entropy-weighted hybrid pooling in CNN for ultrasound breast cancer detection

  • by Ratapong Onjun, Papon Tantiwanichanon, Songkiat Lowmunkhong, Tanakorn Sritarapipat, Sayan Kaennakham, Niwatchai Namwichaisirikul, Kitirat Phattaramarut

  • July 8, 2026

Related Content

ASCO26: Sara Tolaney, MD, MPH, ASCENT-03

Dana-Farber Cancer Institute's Dr. Sara Tolaney from the ASCENT-03 stud in breast cancer. The data further strengthens the case for approval of sacituzumab govitecan as first-line treatment for patients with advanced triple-negative breast cancer.

  • July 1, 2026

  • 1 min
The Unrealized Potential of Cascade Testing

Large language models for breast cancer treatment planning: a blinded real-world evaluation of DeepSeek, ChatGPT, and oncologist recommendations