Health system-scale language models are all-purpose prediction engines

Physicians make critical time-constrained decisions every day. Clinical predictive models can help physicians and administrators make decisions by forecasting clinical and operational events. Existing structured data-based clinical predictive models have limited use in everyday practice owing to complexity in data processing, as well as model development and deployment1-3. Here we show that unstructured clinical notes from the electronic health record can enable the training of clinical language models, which can be used as all-purpose clinical predictive engines with low-resistance development and deployment. Our approach leverages recent advances in natural language processing4,5 to train a large language model for medical language (NYUTron) and subsequently fine-tune it across a wide range of clinical and operational predictive tasks. We evaluated our approach within our health system for five such tasks: 30-day all-cause readmission prediction, in-hospital mortality prediction, comorbidity index prediction, length of stay prediction, and insurance denial prediction. We show that NYUTron has an area under the curve (AUC) of 78.7-94.9%, with an improvement of 5.36-14.7% in the AUC compared with traditional models. We additionally demonstrate the benefits of pretraining with clinical text, the potential for increasing generalizability to different sites through fine-tuning and the full deployment of our system in a prospective, single-arm trial. These results show the potential for using clinical language models in medicine to read alongside physicians and provide guidance at the point of care.

Federated Learning used for predicting outcomes in SARS-COV-2 patients

'Federated Learning' (FL) is a method to train Artificial Intelligence (AI) models with data from multiple sources while maintaining anonymity of the data thus removing many barriers to data sharing. During the SARS-COV-2 pandemic, 20 institutes collaborated on a healthcare FL study to predict future oxygen requirements of infected patients using inputs of vital signs, laboratory data, and chest x-rays, constituting the "EXAM" (EMR CXR AI Model) model. EXAM achieved an average Area Under the Curve (AUC) of over 0.92, an average improvement of 16%, and a 38% increase in generalisability over local models. The FL paradigm was successfully applied to facilitate a rapid data science collaboration without data exchange, resulting in a model that generalised across heterogeneous, unharmonized datasets. This provided the broader healthcare community with a validated model to respond to COVID-19 challenges, as well as set the stage for broader use of FL in healthcare.

SINTESIS DAN KARAKTERISASI NUKLEOTIDA BERTANDA [α-32P]ATP

The utilization of nuclear technology in health sector with molecular techniques is increasingly developed today, especially in Indonesia.  One of which is nucleotide compound marked with [α-32P]ATP, this compound has been used as tracer for deoxyribonucleic acid (DNA)/ ribonucleic acid (RNA) in the study of various physiological and pathological processes. [α-32P]ATP is synthesized through several stages of continuous reaction in one reaction vessel. It begins with synthesis of [γ-32P]ATP through an enzymatic reaction, using H332PO4 and ADP, and enzymes  of lactate dehydrogenase, 3-phosphoglycerate phosphokinase and glyceraldehide 3-phospho  dehydrogenase; followed by phosphorilation of 3’AMP with T4 polinucleotide kinase enzyme to produce 3’-[5’-32P]ATP. The result is hydrolyzed with nuclease P1 enzyme to produce [5’-32P]AMP. The unreacted [γ-32P] is degraded by the addition of hexokinase enzyme and glucose. At the final stage of the reaction, the [5’-32P]AMP is  phosphorilated using phosphoenol-piruvat, piruvat kinase, and myokinase to produce [α-32P]ATP. The test results show that the every stage of reaction is characterized using TLC method, PEI cellulose paper as stationary phase and KH3PO4 0,5 M pH 3,5 as mobile phase. At the end of reaction, the yield of [α-32P]ATP reaches 71,7%, at Rf = 0,2.