Automatic method for assessment of proliferation index in digital images of DLBCL tissue section

摘要: Abstract Diffuse large B-cell lymphoma (DLBCL) is a fast-growing and aggressive neoplasm originating from B lymphocytes. Evaluation of proliferation index (PI) based on Ki67 immunohistochemical nuclear staining used to distinguish proliferating (immunopositive) nonproliferating (immunonegative) cells. Human interpretation PI varies time-consuming, therefore automatic computer-assisted approach may facilitate the performance. Herein we propose new fully estimation (FLAPIE) algorithm, dedicated detection immunopositive immunonegative nuclei, evaluation in digital microscopy images DAB&H-stained samples patients with high-grade DLBCL. FLAPIE performs nuclei original RGB colour space independent image brightness due its textural-statistical approach. Validation was performed 61 non-overlapping whole-slide fragments compared results by QuPath open-source software, MetPiKi algorithm manual two observers. Interobserver agreement calculated between count PIs High concordance found both DAB H-stained count, Compared MetPiKi, presented improved detection. In contrary QuPath, all closely matched number DAB-stained evaluated No significant difference computational methods achieved good prospectively aims serve as tool for clinical application support selection decision treatment.