Low responsiveness of machine learning models to critical or deteriorating health conditions.

Background: Machine learning (ML) based mortality prediction models can be immensely useful in intensive care units. Such a model should generate warnings to alert physicians when a patient's condition rapidly deteriorates, or their vitals are in highly abnormal ranges. Before clinical deployment, it is important to comprehensively assess a model's ability to recognize critical patient conditions.

Epigenomic tomography for probing spatially defined chromatin state in the brain.

Spatially resolved epigenomic profiling is critical for understanding biology in the mammalian brain. Single-cell spatial epigenomic assays were developed recently for this purpose, but they remain costly and labor intensive for examining brain tissues across substantial dimensions and surveying a collection of brain samples. Here, we demonstrate an approach, epigenomic tomography, that maps spatial epigenomes of mouse brain at the scale of centimeters.

Subpopulation-specific machine learning prognosis for underrepresented patients with double prioritized bias correction.

Background: Many clinical datasets are intrinsically imbalanced, dominated by overwhelming majority groups. Off-the-shelf machine learning models that optimize the prognosis of majority patient types (e.g.