CHOICE OF RENAL REPLACEMENT THERAPY IN PATIENTS WITH DIABETIC END STAGE RENAL DISEASE

Attention2majority: Weak multiple instance learning for regenerative kidney grading on whole slide images

Deep learning consistently demonstrates high performance in classifying and segmenting medical images like CT, PET, and MRI. However, compared to these kinds of images, whole slide images (WSIs) of stained tissue sections are huge and thus much less efficient to process, especially for deep learning algorithms. To overcome these challenges, we present attention2majority, a weak multiple instance learning model to automatically and efficiently process WSIs for classification. Our method initially assigns exhaustively sampled label-free patches with the label of the respective WSIs and trains a convolutional neural network to perform patch-wise classification. Then, an intelligent sampling method is performed in which patches with high confidence are collected to form weak representations of WSIs. Lastly, we apply a multi-head attention-based multiple instance learning model to do slide-level classification based on high-confidence patches (intelligently sampled patches). Attention2majority was trained and tested on classifying the quality of 127 WSIs (of regenerated kidney sections) into three categories. On average, attention2majority resulted in 97.4%±2.4 AUC for the four-fold cross-validation. We demonstrate that the intelligent sampling module within attention2majority is superior to the current state-of-the-art random sampling method. Furthermore, we show that the replacement of random sampling with intelligent sampling in attention2majority results in its performance boost (from 94.9%±3.1 to 97.4%±2.4 average AUC for the four-fold cross-validation). We also tested a variation of attention2majority on the famous Camelyon16 dataset, which resulted in 89.1%±0.8 AUC¹. When compared to random sampling, the attention2majority demonstrated excellent slide-level interpretability. It also provided an efficient framework to arrive at a multi-class slide-level prediction.

Comparison of survival between diabetic and non-diabetic patients on maintenance hemodialysis: a single-centre experience

We analyzed survival rates of 144 prevalent patients on maintenance hemodialysis from 1998 to 2003 at the Department of Nephrology and Dialysis, Rijeka University Hospital, Rijeka, Croatia, and evaluated risk factors predicting their survival. Included were only end-stage renal disease (ESRD) patients on maintenance hemodialysis treatment dialysed more than 6 months before entering the study and who were clinically stable. The patients were randomised in two groups according to the presence or absence of diabetic nephropathy as the cause of ESRD and followed-up. The patient's death as outcome measure was recorded. The survival rates were estimated by the Kaplan-Meier method. The major causes of death were cardiovascular disease in 40 (60.6%) patients. An acute myocardial infarction in 15 (22.7%) patients was the major single cause of death. We found a significantly lower survival of diabetic patients than non-diabetic patients (P=0.0013). The most important predictors of death among diabetic patients on maintenance hemodialysis were hyperglycaemia (P<0.001), ischemic heart disease (P=0.004), hypercholesterolemia (P=0.013), and low delivered dialysis dose (P=0.013). The survival of diabetic patients undergoing hemodialysis was much worse than survival of non-diabetic patients. The cardiovascular disease remained the major cause of death in both groups. Early detection of pre-existing cardiovascular risk factors and diseases, and treatment of infections leading to sepsis, are of great importance, as they may influence the survival rates. Intensive management of diabetic patients is essential.