ABSTRACT
AIMS: In routine diagnosis of lymphoma, initial non-specialist triage is carried out when the sample is biopsied to determine if referral to specialised haematopathology services is needed. This places a heavy burden on pathology services, causes delays and often results in over-referral of benign cases. We aimed to develop an automated triage system using artificial intelligence (AI) to enable more accurate and rapid referral of cases, thereby addressing these issues. METHODS: A retrospective dataset of H&E-stained whole slide images (WSI) of lymph nodes was taken from Newcastle University Hospital (302 cases) and Manchester Royal Infirmary Hospital (339 cases) with approximately equal representation of the 3 most prevalent lymphoma subtypes: follicular lymphoma, diffuse large B-cell and classic Hodgkin's lymphoma, as well as reactive controls. A subset (80%) of the data was used for training, a further validation subset (10%) for model selection and a final non-overlapping test subset (10%) for clinical evaluation. RESULTS: AI triage achieved multiclass accuracy of 0.828±0.041 and overall accuracy of 0.932±0.024 when discriminating between reactive and malignant cases. Its ability to detect lymphoma was equivalent to that of two haematopathologists (0.925, 0.950) and higher than a non-specialist pathologist (0.75) repeating the same task. To aid explainability, the AI tool also provides uncertainty estimation and attention heatmaps. CONCLUSIONS: Automated triage using AI holds great promise in contributing to the accurate and timely diagnosis of lymphoma, ultimately benefiting patient care and outcomes.
ABSTRACT
CONTEXT.: Women with diabetes have increased stillbirth risk. Although the underlying pathophysiological processes are poorly understood, stillbirth is frequently related to abnormal placental structure and function. OBJECTIVE.: To investigate placental morphology and cellular characteristics in the placentas of women with diabetes who had stillbirths and stillbirths of unexplained cause. DESIGN.: Placentas from women with uncomplicated live births, live births in women with diabetes, unexplained stillbirths, and stillbirths related to diabetes (n = 10/group) underwent clinical histopathologic assessment and were also investigated using immunohistochemical staining to quantify syncytial nuclear aggregates, proliferation, trophoblast area, vascularization, T cells, placental macrophages (Hofbauer cells), and the receptor for advanced glycation end products. RESULTS.: Ki67+ cells were decreased in unexplained stillbirths compared with live births in women with diabetes. Both stillbirth groups had increased cytokeratin 7+/nuclear area compared with controls. Blood vessels/villi were decreased in unexplained stillbirth compared with live births from women with diabetes. Compared with uncomplicated controls, CD163+ macrophages were increased in live births in women with diabetes and unexplained stillbirths, and further increased in stillbirths related to diabetes. There was no change in CD3+ T cells or syncytial nuclear aggregates. Receptor for advanced glycation end products-positive cells were decreased in both stillbirth groups compared with diabetes-related live births. Co-localization of receptor for advanced glycation end products in macrophages was increased in both stillbirth groups compared with live birth groups. CONCLUSIONS.: Stillbirths related to diabetes exhibit placental phenotypic differences compared with live births. Further investigation of these parameters may provide understanding of the pathologic mechanisms of stillbirth and aid the development of stillbirth prevention strategies.
