Can kidney disease be seen through the eyes? High-tech doctor's "AI fortune telling" 2 system works: Knows the hypotension problem of kidney dialysis patients in advance
Can kidney disease be seen through the eyes? Doctor’s “AI fortune telling” 2 system works
How can you see the kidneys from the eyes? How to create safe and high-quality medical care for kidney dialysis patients? “Kaohsiung Joint Learning Smart Healthcare Alliance” has established the world’s first generative AI medical industry cluster in Kaohsiung’s Asia New Bay Area. Generative AI has advantages in improving medical efficiency and doctor-patient relationships, and has automatic medical record summary and paper reading summary. functions.
The Chung Ho Memorial Hospital affiliated to Kaohsiung Medical University has created an AI prediction system for chronic disease lesions. Through the “Diabetic Nephropathy and Exacerbation Risk Prediction System”, doctors can see the kidneys from the eyes and predict the nephropathy of diabetic patients through diabetic fundus photography; " “Prediction System for Predicting Hypotension and Comorbidities in Dialysis Patients” can provide early warning of complications that may occur in dialysis patients during treatment, allowing medical staff to assess in advance whether dialysis parameters need to be adjusted to create high-quality kidney dialysis.
Dr. Huang Shangzhi, deputy director of the Chung Ho Memorial Hospital affiliated to Kaohsiung Medical University, pointed out that in an era of rapid technological change, the development of medical technology is constantly innovating. In particular, ChatGPT has begun to use generative AI to drive the global artificial intelligence generative technology competition. The medical system of Gaoyei has been committed to promoting smart medical care in recent years. The Chunghwa Memorial Hospital affiliated to Gaoyei is a university hospital and medical center, and it has also actively invested in the research and development of smart health care. The results this year (2023) are also quite fruitful.
Diabetic nephropathy prediction system can predict the kidney disease problems of diabetic patients in advance
This time, the alliance of high-tech doctors participated in the research and development of two chronic disease prediction systems, “prediction system for predicting hypotension and comorbidities in dialysis patients” and “prediction system for diabetic nephropathy and exacerbation risk”, making AI a powerful assistant for doctors and assisting patients. Diagnosis, treatment and risk prediction. And through the Kaohsiung 5G “AIoT Smart Medical and Cross-hospital Joint Verification and Validation Platform”, the platform results will be used to implement the validation mechanism of software medical materials, accelerate the industrialization of software medical materials SaMD, and drive the development of Kaohsiung’s smart medical industry and the clinical application of AI .
Dr. Wu Bingxun from the Department of Nephrology of Gaoyin Medical University said that diabetic fundus lesions and diabetic kidney disease are both small blood vessel complications, and the two are closely related. Gaoyi’s “Diabetic Nephropathy and Exacerbation Risk Prediction System” uses AI deep learning artificial intelligence methods to conduct type-training based on images of diabetic eye lesions and clinical biochemical test data and statistically analyze the correlation and linkage.
It is hoped that clinical medical staff can use the model to predict the patient’s risk of kidney disease and intervene early to delay the occurrence of kidney dialysis, and truly implement preventive medicine that combines AI big data analysis and integration to make medical treatment more accurate and reduce medical expenses; in the future, it will Introduce it into general health check-up services or community health education to reduce the chance of disease and complications.
Predicting comorbidities in dialysis patients to avoid “hypotension” and potential death risks
Director Qiu Yiwen of the Hematology and Peritoneal Dialysis Unit of Gaoyi Medical University explained that since hemodialysis (kidney dialysis) patients require 4 hours of treatment three times a week, the amount of blood removed from the patient’s body during the dialysis process is as high as 250 ml per minute. The water removed per minute is between 5-20 ml, which can easily cause hypotension, thereby increasing the risk of death for dialysis patients.
The “prediction system for predicting hypotension and comorbidities in dialysis patients” can provide early warning of complications that may occur during treatment, allowing medical staff to assess in advance whether dialysis parameters need to be adjusted, improving the quality of kidney dialysis patients and making kidney dialysis more efficient. Safety. In the future, the predicted hourly blood pressure values will be combined to evaluate the possible causes of complications and carry out treatment based on the prediction results.
The Chung Ho Memorial Hospital affiliated to Kaohsiung Medical University cooperates with the Taiwan Artificial Intelligence Laboratory and other medical systems in Kaohsiung to train smart medical products that can share results through decentralized joint learning technology, combining the Kaohsiung Software Park, ARVR, and Science Park Together with the metal center’s advantages in smart medical material production, Kaohsiung will become a rapid production and manufacturing center for smart medical care in Asia and even the world, and together we will create “the first in the world.”
Further reading: Why does aged steak taste better? Taiwan Food and Drug Administration’s answer: “1 process” improves tenderness and juiciness
