| Autor: |
Chen YC; Division of Digital Pathology, Department of Anatomical Pathology, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Pathology, School of Medicine, Tzu Chi University, Hualien 970, Taiwan.; Institute of Medical Sciences, Tzu Chi University, Hualien 970, Taiwan., Lin SZ; Institute of Medical Sciences, Tzu Chi University, Hualien 970, Taiwan.; Bioinnovation Center, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Neuroscience Center, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Neurosurgery, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Surgery, School of Medicine, Tzu Chi University, Hualien 970, Taiwan., Wu JR; Integration Center of Traditional Chinese and Modern Medicine, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Medical Research, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan., Yu WH; aetherAI, Co., Ltd., Taipei 115, Taiwan., Harn HJ; Bioinnovation Center, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Division of Molecular Pathology, Department of Anatomical Pathology, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan., Tsai WC; Department of Pathology, Tri-Service General Hospital, National Defense Medical Center, Taipei 114, Taiwan., Liu CA; Bioinnovation Center, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Neuroscience Center, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Medical Research, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan., Kuo KL; YuanLi Instrument Co., Ltd., Taipei 114, Taiwan., Yeh CY; aetherAI, Co., Ltd., Taipei 115, Taiwan., Tsai ST; Institute of Medical Sciences, Tzu Chi University, Hualien 970, Taiwan.; Department of Neuroscience Center, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Neurosurgery, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien 970, Taiwan.; Department of Surgery, School of Medicine, Tzu Chi University, Hualien 970, Taiwan. |
| Abstrakt: |
Interobserver variations in the pathology of common astrocytic tumors impact diagnosis and subsequent treatment decisions. This study leveraged a residual neural network-50 (ResNet-50) in digital pathological images of diffuse astrocytoma, anaplastic astrocytoma, and glioblastoma to recognize characteristic pathological features and perform classification at the patch and case levels with identification of incorrect predictions. In addition, cellularity and nuclear morphological features, including axis ratio, circularity, entropy, area, irregularity, and perimeter, were quantified via a hybrid task cascade (HTC) framework and compared between different characteristic pathological features with importance weighting. A total of 95 cases, including 15 cases of diffuse astrocytoma, 11 cases of anaplastic astrocytoma, and 69 cases of glioblastoma, were collected in Taiwan Hualien Tzu Chi Hospital from January 2000 to December 2021. The results revealed that an optimized ResNet-50 model could recognize characteristic pathological features at the patch level and assist in diagnosis at the case level with accuracies of 0.916 and 0.846, respectively. Incorrect predictions were mainly due to indistinguishable morphologic overlap between anaplastic astrocytoma and glioblastoma tumor cell area, zones of scant vascular lumen with compact endothelial cells in the glioblastoma microvascular proliferation area mimicking the glioblastoma tumor cell area, and certain regions in diffuse astrocytoma with too low cellularity being misrecognized as the glioblastoma necrosis area. Significant differences were observed in cellularity and each nuclear morphological feature among different characteristic pathological features. Furthermore, using the extreme gradient boosting (XGBoost) algorithm, we found that entropy was the most important feature for classification, followed by cellularity, area, circularity, axis ratio, perimeter, and irregularity. Identifying incorrect predictions provided valuable feedback to machine learning design to further enhance accuracy and reduce errors in classification. Moreover, quantifying cellularity and nuclear morphological features with importance weighting provided the basis for developing an innovative scoring system to achieve objective classification and precision diagnosis among common astrocytic tumors. |
