Infant digestive health assessment and early warning system

Abstract

A neonatal digestive disease auxiliary diagnosis and health early warning system is disclosed to address the problem of misdiagnosis when using color charts to compare infant excrement colors in assessing bile retention. The system includes: an image acquisition unit that captures an image of a newborn's excrement; a recognition module that trains a deep learning recognition model using several images of the newborn's excrement to identify the color pattern of the excrement image as normal or abnormal; a classification module that maps the color pattern to a reference color number on a fecal color chart and generates a color risk classification result; and a calculation module that receives total bilirubin and direct bilirubin test data from a newborn, performs ratio calculations and unit conversions, and generates bile retention risk warning information. This system assists in the diagnosis and provides early warning of neonatal digestive diseases.

Claims

1. A neonatal digestive disease auxiliary judgment and health early warning system, comprising: an image acquisition unit for capturing an image of a newborn's excrement; a recognition module coupled to the image acquisition unit, the recognition module having a deep learning recognition model, the deep learning recognition model being pre-trained with deep learning and identifying the color pattern of the newborn's excrement image as normal or abnormal under various environmental conditions; a classification module coupled to the recognition module, mapping the color pattern result identified by the recognition module to a reference color number of a fecal color card, and generating a color risk classification result; a calculation module receiving detection data of total bilirubin and direct bilirubin of a newborn, the calculation module performing ratio calculation and unit conversion, and generating bile retention risk warning information; and an early warning module coupled to the classification module and the calculation module, the early warning module generating an immediate warning and providing reference information for subsequent actions related to medical assessment or medical consultation based on the color risk classification result and the bile retention risk warning information.

2. As in request item 1, a neonatal digestive disease auxiliary judgment and health early warning system, wherein, The image capturing unit performs image preprocessing on the newborn's excrement image, including cropping, brightness adjustment, and color correction.

3. As in request item 1, a neonatal digestive disease auxiliary judgment and health early warning system, wherein, The deep learning recognition model of the recognition module automatically performs multi-level feature analysis, and the recognition module performs closed-loop recognition according to the preset workflow.

4. As in request item 1, a neonatal digestive disease auxiliary judgment and health early warning system, wherein, The color card for this type of stool, which is used in this classification module, is a Taiwanese infant stool identification card.

5. As in request item 1, a neonatal digestive disease auxiliary judgment and health early warning system, wherein, The calculation module receives the total bilirubin and direct bilirubin test values ​​of the newborn in various measurement units, and converts them into a standardized unit and calculates the ratio.

6. The neonatal digestive disease auxiliary judgment and health early warning system as described in any of the requests 1 to 5, further includes an update module for periodically updating the deep learning recognition model and the health-related reference information used therein.

7. The neonatal digestive disease auxiliary judgment and health early warning system as described in any of the requests 1 to 5, further includes an interactive module, through which a user switches the shooting function of the image capturing unit and the conversion function of the calculation module, and inputs the detection values ​​of the newborn's total bilirubin and direct bilirubin through the interactive module.

8. As in request item 7, a neonatal digestive disease auxiliary judgment and health early warning system, wherein, The interactive module has a display interface that uses images and text to show the switching of the operation functions of the recognition module and the calculation module. It displays the image of the newborn's excrement captured by the image capturing unit, and integrates the inference confidence value of the deep learning recognition model of the recognition module, the color risk classification result of the classification module, and the real-time warning and reference information related to medical assessment or medical consultation of the early warning module.

Images