Rxivist logo

Deep Learning-Based Universal Expert-Level Recognizing Pathological Images of Hepatocellular Carcinoma and beyond

By Wei-Ming Chen, Min Fu, Cheng-Ju Zhang, Qing-Qing Xing, Fei Zhou, Meng-Jie Lin, Xuan Dong, Qi-Zhong Zheng, Mei-Zhu Hong, Jin-Shui Pan

Posted 27 Mar 2020
medRxiv DOI: 10.1101/2020.03.22.20041178

PurposeHuman-based medical-image interpretation always falls into the predicament between specialized practitioners and expanding medical imaging. We aim at developing a diagnostic tool for pathological-image classification by using transfer learning that can be applied to diverse tumor types. Experimental DesignIn this study, images were retrospectively collected and prospectively analyzed using machine learning. Microscopic images of liver tissue that show or do not show hepatocellular carcinoma were used to train and validate a classification framework based on convolutional neural network. To evaluate the universal classification performance of the artificial-intelligence (AI) framework, histological images from colorectal tissue and breast were also collected. Training and validation set of images were collected from Xiamen Hospital of Traditional Chinese Medicine whereas test set of images were collected from Zhongshan Hospital Xiamen University. ResultsAccuracy, sensitivity, and specificity were reported and compared to human image interpretation and other AI image classification systems such as AlexNet and GoogLeNet. For the test dataset, sensitivity, specificity, and area under the curve of the AI framework were 99.1%, 98.0%, and 0.960, respectively. In human-machine comparisons, the accuracy of the AI framework was 98.5%, while the accuracy of human experts fluctuated between 93.0% and 95.0%. Based on transfer learning, the AI framework accuracy for colorectal carcinoma, breast invasive ductal carcinoma, were 96.8%, and 96.0%, respectively. ConclusionsThe performance of the proposed AI framework in classifying histological images with hepatocellular carcinoma is comparable to the classification by human experts. With limited training, the proposed AI framework has potential universality in histological image classification. Study Highlights O_TEXTBOXWHAT IS KNOWN{checkmark} Accurate recognition of medical images is the basis for clinical decision-making. {checkmark}Unresolved challenge exists between specialized practitioners and expanding medical imaging output. WHAT IS NEW HERE{checkmark} Proposed AI framework has excellent performance in classifying hepatocellular carcinoma. {checkmark}The AI framework has universal feature in classifying other types of medical images. {checkmark}The AI framework may help to interpret emerging imaging technology. C_TEXTBOX

Download data

  • Downloaded 663 times
  • Download rankings, all-time:
    • Site-wide: 51,450
    • In pathology: 239
  • Year to date:
    • Site-wide: 66,397
  • Since beginning of last month:
    • Site-wide: 36,392

Altmetric data


Downloads over time

Distribution of downloads per paper, site-wide


PanLingua

News