Immediate postnatal prediction of death or bronchopulmonary dysplasia among very preterm and very low birth weight infants based on gradient boosting decision trees algorithm: A nationwide database study in Japan.

INTRODUCTION: Bronchopulmonary dysplasia (BPD) poses a substantial global health burden. Individualized treatment strategies based on early prediction of the development of BPD can mitigate preterm birth complications; however, previously suggested predictive models lack early postnatal applicability. We aimed to develop predictive models for BPD and mortality based on immediate postnatal clinical data. METHODS: Clinical information on very preterm and very low birth weight infants born between 2008 and 2018 was extracted from a nationwide Japanese database. The gradient boosting decision trees (GBDT) algorithm was adopted to predict BPD and mortality, using predictors within the first 6 h postpartum. We assessed the temporal validity and evaluated model adequacy using Shapley additive explanations (SHAP) values. RESULTS: We developed three predictive models using data from 39,488, 39,096, and 40,291 infants to predict "death or BPD," "death or severe BPD," and "death before discharge," respectively. These well-calibrated models achieved areas under the receiver operating characteristic curve of 0.828 (95% CI: 0.828-0.828), 0.873 (0.873-0.873), and 0.887 (0.887-0.888), respectively, outperforming the multivariable logistic regression models. SHAP value analysis identified predictors of BPD, including gestational age, size at birth, male sex, and persistent pulmonary hypertension. In SHAP value-based case clustering, the "death or BPD" prediction model stratified infants by gestational age and persistent pulmonary hypertension, whereas the other models for "death or severe BPD" and "death before discharge" commonly formed clusters of low mortality, extreme prematurity, low Apgar scores, and persistent pulmonary hypertension of the newborn. CONCLUSIONS: GBDT models for predicting BPD and mortality, designed for use within 6 h postpartum, demonstrated superior prognostic performance. SHAP value-based clustering, a data-driven approach, formed clusters of clinical relevance. These findings suggest the efficacy of a GBDT algorithm for the early postnatal prediction of BPD.

Assessment of patients' characteristics associated with the efficacy and safety of oral valganciclovir treatment for infants with symptomatic congenital cytomegalovirus disease.

INTRODUCTION: Insurance coverage for oral valganciclovir (VGCV) began in Japan in April 2023 on the basis of results, including our clinical trials for symptomatic congenital cytomegalovirus (CMV) disease. The VGCV treatment is available throughout Japan, so clinicians must consider the likelihood of hearing improvement and the possibility of neutropenia before dosing. MATERIALS AND METHODS: We performed a substudy of an investigator-initiated, single-arm, prospective, multicenter, clinical trial in which 24 infants with symptomatic congenital CMV disease were orally administered 16 mg/kg VGCV twice daily for 6 months as an intervention. We examined the infants' baseline characteristics associated with improved hearing impairment or a severely reduced neutrophil count. RESULTS: Of the 24 patients, 4 had normal hearing on assessment of their ear with the best hearing. Hearing impairment improved in 14 patients and did not respond to VGCV treatment in 6 patients at the 6-month hearing assessment. CMV DNA levels in plasma at baseline were higher in patients in whom hearing did not respond to treatment. A neutrophil count <500/mm3 occurred in 5 (21%) patients for the first 6 weeks and in 8 (33%) patients for the first 6 months. A neutrophil count at screening and the lowest neutrophil count over the 6 months showed the highest correlation (r = 0.477, p = 0.019). CONCLUSIONS: Infants with a low plasma viral load at screening tend to have an improvement in hearing impairment. Clinicians should be aware of neutropenia during VGCV treatment particularly in patients with a low neutrophil count during screening.

Image diagnosis models for the oral assessment of older people using convolutional neural networks: A retrospective observational study.

AIMS: The purpose of this study was to construct a model for oral assessment using deep learning image recognition technology and to verify its accuracy. BACKGROUND: The effects of oral care on older people are significant, and the Oral Assessment Guide has been used internationally as an effective oral assessment tool in clinical practice. However, additional training, education, development of user manuals and continuous support from a dental hygienist are needed to improve the inter-rater reliability of the Oral Assessment Guide. DESIGN: A retrospective observational study. METHODS: A total of 3,201 oral images of 114 older people aged >65 years were collected from five dental-related facilities. These images were divided into six categories (lips, tongue, saliva, mucosa, gingiva, and teeth or dentures) that were evaluated by images, out of the total eight items that comprise components of the Oral Assessment Guide. Each item was classified into a rating of 1, 2 or 3. A convolutional neural network, which is a deep learning method used for image recognition, was used to construct the image recognition model. The study methods comply with the STROBE checklist. RESULTS: We constructed models with a classification accuracy of 98.8% for lips, 94.3% for tongue, 92.8% for saliva, 78.6% for mucous membranes, 93.0% for gingiva and 93.6% for teeth or dentures. CONCLUSIONS: Highly accurate diagnostic imaging models using convolutional neural networks were constructed for six items of the Oral Assessment Guide and validated. In particular, for the five items of lips, tongue, saliva, gingiva, and teeth or dentures, models with a high accuracy of over 90% were obtained. RELEVANCE TO CLINICAL PRACTICE: The model built in this study has the potential to contribute to obtain reproducibility and reliability of the ratings, to shorten the time for assessment, to collaborate with dental professionals and to be used as an educational tool.

Quantitative assessment of visual estimation of the infrared indocyanine green imaging of lymph nodes retrieved at sentinel node navigation surgery for gastric cancer.

BACKGROUND: Although the infrared indocyanine green (ICG) imaging is an effective method to identify sentinel lymph nodes (SLNs) of gastric cancer, its objectivity has not been verified. METHODS: We studied 563 lymph nodes under infrared light observation from the ICG-positive lymphatic basins of 36 patients who underwent SLN-navigated gastrectomy for clinically node-negative gastric cancer. First, the rate of SLN detection, the number of SLNs and sensitivities were compared between ordinary light observation and infrared light observation. Second, 563 lymph nodes were grouped into ICG-positive and -negative under infrared light observation. The intensities of the region of interest for each lymph node defined as the lymph node on which digital imaging was performed using an imaging-software, and the region of reference defined as its surrounding background, were compared and quantified. RESULTS: In the comparison of ordinary light observation with infrared light observation, the SLN identification rates were 28/36 (78 %) vs. 36/36 (100 %), the mean ± SD (minimum to maximum) number of SLNs was 3.4 ± 3.7 (0-16) vs. 9.2 ± 5.9 (2-25), and the sensitivities were 1/5 (20 %) vs. 5/5 (100 %). The ICG-positive group contained 358 lymph nodes with an intensity of 0.323 ± 1.56 (mean ± SD), and the ICG-negative group contained 205 lymph nodes with an intensity of 0.639 ± 1.93 (mean ± SD), demonstrating a significant difference between these two groups (P < 0.0001). CONCLUSIONS: The significant difference in the intensity as measured by an imaging-software between ICG-positive and ICG-negative lymph nodes would erase the concern about the objectivity of the infrared ICG method for SLN-navigated surgery for early gastric cancer.

Assessment of left ventricular dyssynchrony during development of heart failure by a novel program using ECG-gated myocardial perfusion SPECT.

BACKGROUND: A novel program, "cardioGRAF", has been developed to analyze regional left ventricular (LV) systolic/diastolic function and dyssynchrony, so the present study aimed to use it confirm the presence of LV dyssynchrony, and to correlate LV function and dyssynchrony with plasma B-type natriuretic peptide (BNP) levels during the early to advanced stages of heart failure (HF). METHODS AND RESULTS: Fourteen control subjects (G-C) and 50 patients (New York Heart Association functional class I: G-1, 21 patients; class II: G-2, 15 patients; and class III: G-3, 14 patients) were examined by ECG-gated myocardial perfusion single-photon emission computed tomography, using the new index of dyssynchrony, maximal difference (MD), which is the difference between the earliest and latest temporal parameters among 17 segments. First-third filling rate (FR) and the MD of time to peak FR revealing diastolic dyssynchrony were significantly different between G-C subjects and G-1 patients. Ejection fraction, peak ejection rate, peak FR, MD of time to end-systole, and MD of time to peak ejection rate were significantly correlated with plasma BNP levels. CONCLUSION: Diastolic dyssynchrony was demonstrated even in the early stage of HF, but, although not correlated with the plasma BNP level, systolic dyssynchrony might affect it.

Cerebral ventricular volume assessment by three-dimensional ultrasonography.

We assessed ventricular volume with three-dimensional ultrasonography. The study group consisted of 54 infants admitted to the neonatal intensive care unit (NICU). Gestational age was 32.9+/-3.5 weeks and birth weight was 1774+/-623 g. To obtain images, the transducer was placed on the anterior fontanel. Sagittal sections were serially and automatically scanned from left to right. The angle of scanning was 60 degrees and five seconds were spent on e scanning. Two hundred fifty B-mode images were stored in the built-in computer for later retrieval. We chose the parasagittal sections from the midline to the lateral portion and traced manually an area of the lateral ventricle in each section at intervals of 1 mm. The built-in computer integrated them and measured the ventricular volume. The lateral ventricular size became larger during the first two weeks after birth. The left ventricle was larger than the right one. There was no correlation between lateral ventricular volume and birth weight. Our study shows the normal value of the lateral ventricular volume in neonates.