Selected Publications

We develop a mortality prediction algorithm which outperforms existing methods, and apply transfer learning to overcome the barrier of data scarcity during site implementation.
In Biomedical Informatics Insights

Sepsis can be predicted hours in advance of onset, using only vital signs, with better performance than methods in standard practice today. High-order correlations between vital signs are key.
In Computers in Biology and Medicine

We develop a rapid mix-and-measure method for identifying and classifying potential G-quadruplex-forming sequences in the human genome.
In NAR

Our results suggest telomeric overhang length and dynamics may contribute to the regulation of telomere extension via telomerase action and the ALT mechanism.
In Structure

Recent Publications

More Publications

  • Using transfer learning for improved mortality prediction in a data-scarce hospital setting

    Details PDF

  • Cost and mortality impact of an algorithm-driven sepsis prediction system

    Details PDF

  • Using electronic health record collected clinical variables to predict medical intensive care unit mortality

    Details PDF

  • Prediction of sepsis in the intensive care unit with minimal electronic health record data: a machine learning approach

    Details PDF

  • A computational approach to mortality prediction of alcohol use disorder inpatients

    Details PDF

  • A computational approach to early sepsis detection

    Details PDF

  • High-performance detection and early prediction of septic shock for alcohol-use disorder patients

    Details PDF

  • Quantitative analysis and prediction of G-quadruplex forming sequences in double-stranded DNA

    Details PDF

  • Discharge recommendation based on a novel technique of homeostatic analysis

    Details PDF

  • G-quadruplex formation in double strand DNA probed by NMM and CV fluorescence

    Details PDF

Contact