Search

Simple Summary: We investigated a new method for diagnosing and predicting outcomes in canine pulmonary carcinoma. We developed a deep learning-based algorithm that accurately detects tumor nuclei and subsequently measures size and shape parameters. The variation in nuclear size and shape (nuclear pleomorphism) is a crucial malignancy criterion used in the current grading system for canine pulmonary carcinoma. Pathologists currently evaluate it and classify it according to a three-tier system. Manual morphometry is a more objective approach where tumor nuclei are individually encircled and analyzed. This task can be easily performed by an algorithm. Our algorithm's accuracy in correctly detecting and segmenting tumor nuclei was considered good when compared to manual morphometry. By comparing automated morphometry with conventional prognostic tests, such as pathologists' estimates, mitotic count, histological grading, and clinical staging, we found that our approach was equally accurate in terms of prognostic value. The algorithm's advantage lies in its high reproducibility and efficiency. Automated evaluation of nuclear pleomorphism can enhance the efficiency and reliability of canine pulmonary carcinoma diagnosis and grading, effectively addressing issues of inter-observer reproducibility. However, further optimization of the algorithm and validation with a larger study group is necessary to confirm our findings. The integration of deep learning-based tools into diagnostic workflows is increasingly prevalent due to their efficiency and reproducibility in various settings. We investigated the utility of automated nuclear morphometry for assessing nuclear pleomorphism (NP), a criterion of malignancy in the current grading system in canine pulmonary carcinoma (cPC), and its prognostic implications. We developed a deep learning-based algorithm for evaluating NP (variation in size, i.e., anisokaryosis and/or shape) using a segmentation model. Its performance was evaluated on 46 cPC cases with comprehensive follow-up data regarding its accuracy in nuclear segmentation and its prognostic ability. Its assessment of NP was compared to manual morphometry and established prognostic tests (pathologists' NP estimates (n = 11), mitotic count, histological grading, and TNM-stage). The standard deviation (SD) of the nuclear area, indicative of anisokaryosis, exhibited good discriminatory ability for tumor-specific survival, with an area under the curve (AUC) of 0.80 and a hazard ratio (HR) of 3.38. The algorithm achieved values comparable to manual morphometry. In contrast, the pathologists' estimates of anisokaryosis resulted in HR values ranging from 0.86 to 34.8, with slight inter-observer reproducibility (k = 0.204). Other conventional tests had no significant prognostic value in our study cohort. Fully automated morphometry promises a time-efficient and reproducible assessment of NP with a high prognostic value. Further refinement of the algorithm, particularly to address undersegmentation, and application to a larger study population are required. [ABSTRACT FROM AUTHOR]