Lung neuroendocrine neoplasms (lung NENs) are categorised by morphology, defining a classification sometimes unable to reflect ultimate clinical outcome. Subjectivity and poor reproducibility characterise diagnosis and prognosis assessment of all NENs. Here, we propose a machine learning framework for tumour prognosis assessment based on a quantitative, automated and repeatable evaluation of the spatial distribution of cells immunohistochemically positive for the proliferation marker Ki-67, performed on the entire extent of high-resolution whole slide images. Combining features from the fields of graph theory, fractality analysis, stochastic geometry and information theory, we describe the topology of replicating cells and predict prognosis in a histology-independent way. We demonstrate how our approach outperforms the well-recognised prognostic role of Ki-67 Labelling Index on a multi-centre dataset comprising the most controversial lung NENs. Moreover, we show that our system identifies arrangement patterns in the cells positive for Ki-67 that appear independently of tumour subtyping. Strikingly, the subset of these features whose presence is also independent of the value of the Labelling Index and the density of Ki-67-positive cells prove to be especially relevant in discerning prognostic classes. These findings disclose a possible path for the future of grading and classification of NENs.

The development of atrial fibrillation (AF) during hospitalization for acute myocardial infarction (AMI) is common and associated with worse prognosis. We aimed to train and, for the first time, cross-nationally validate a machine learning model to predict AF developing during the cardiac intensive care unit (CICU) stay in post-AMI patients admitted in sinus rhythm, using only routine, non-electrocardiographic clinical parameters available at CICU admission. We developed and tested classification pipelines on two independent cohorts of post-AMI patients from Italy (ITA, Sinus Rhythm = 2,200, AF = 240) and Finland (FIN, Sinus Rhythm = 3,631, AF = 452). Models were first evaluated via nested cross-validation on held out samples of the training cohort (internal testing) and then cross-cohort (external testing). An L2-regularized logistic regression model performed optimally in each case, achieving comparable performances in internal (AUROC: 0.740 ITA, 0.755 FIN) and external (AUROC: 0.735 ITA-on-FIN, 0.723 FIN-on-ITA) testing. A combined-cohort model achieved an AUROC of 0.749. A posteriori feature relevance analysis confirmed the predictive value of established risk factors including advanced age, lower left ventricular ejection fraction and hypertension in both cohorts. Ongoing aspirin therapy and non-ST-elevation myocardial infarction emerged as additional protective factors exclusive to cohort ITA, whereas a higher high-sensitivity C-reactive protein (risk factor) and the current smoking status (protective factor) were specific to cohort FIN. Our study demonstrates that machine learning models using routinely collected clinical parameters could help predict in-hospital AF in post-AMI patients across different populations and national healthcare systems. Notably, a simple and fully interpretable logistic regression model achieved robust performance in the first cross-national validation for this prediction task, suggesting that more complex, black-box approaches may offer limited additional benefit for this specific task. Further improvements are needed to reach real-world application in clinical contexts for early risk stratification.