Machine learning for morbid glomerular hypertrophy.

A practical research method integrating data-driven machine learning with conventional model-driven statistics is sought after in medicine. Although glomerular hypertrophy (or a large renal corpuscle) on renal biopsy has pathophysiological implications, it is often misdiagnosed as adaptive/compensatory hypertrophy. Using a generative machine learning method, we aimed to explore the factors associated with a maximal glomerular diameter of ≥ 242.3 µm. Using the frequency-of-usage variable ranking in generative models, we defined the machine learning scores with symbolic regression via genetic programming (SR via GP). We compared important variables selected by SR with those selected by a point-biserial correlation coefficient using multivariable logistic and linear regressions to validate discriminatory ability, goodness-of-fit, and collinearity. Body mass index, complement component C3, serum total protein, arteriolosclerosis, C-reactive protein, and the Oxford E1 score were ranked among the top 10 variables with high machine learning scores using SR via GP, while the estimated glomerular filtration rate was ranked 46 among the 60 variables. In multivariable analyses, the R2 value was higher (0.61 vs. 0.45), and the corrected Akaike Information Criterion value was lower (402.7 vs. 417.2) with variables selected with SR than those selected with point-biserial r. There were two variables with variance inflation factors higher than 5 in those using point-biserial r and none in SR. Data-driven machine learning models may be useful in identifying significant and insignificant correlated factors. Our method may be generalized to other medical research due to the procedural simplicity of using top-ranked variables selected by machine learning.

Accuracy of BIS monitoring using a novel interface device connecting conventional needle-electrodes and BIS sensors during frontal neurosurgical procedures.

BACKGROUND: Bispectral index (BIS) monitoring is a widely used non-invasive method to monitor the depth of anesthesia. However, in the event of surgeries requiring a frontal approach, placement of the electrode may be impossible at the designated area to achieve a proper BIS measurement. METHODS: We developed an investigational interface device to connect needle-electrodes to BIS sensors. The safety and clinical performance were investigated in patients who underwent surgery. Direct BIS values from a disposable BIS electrode and indirect values via the interface device were simultaneously recorded from the same areas of electrode placement in a single patient. The agreement between the direct and indirect BIS values was statistically analyzed. RESULTS: The interface device with a silver electrode demonstrated sufficient electric conduction to transmit electroencephalogram signals. The overall BIS curves were similar to those of direct BIS monitoring. Direct and indirect BIS values from 18 patients were statistically analyzed using a linear mixed model and a significant concordance was confirmed (indirect BIS = 7.0405 + 0.8286 * direct BIS, p<0.0001). Most observed data (2582/2787 data points, 92.64%) had BIS unit differences of 10 or less. CONCLUSIONS: The interface device provides an opportunity for intraoperative BIS monitoring of patients, whose clinical situation does not permit the placement of conventional adhesive sensors at the standard location.