Core Insights - The article discusses a significant advancement in chronic kidney disease (CKD) diagnosis through a non-invasive model utilizing retinal images, developed by a research team from Sun Yat-sen University [2][3][6]. Group 1: Research Development - The Kidney Intelligent Diagnosis System (KIDS) was created to predict kidney biopsy outcomes non-invasively, achieving an area under the curve (AUC) of 0.839-0.993 in CKD screening [3][11]. - KIDS can accurately identify five common pathological types of CKD with an AUC of 0.790-0.932, outperforming nephrologists by 26.98% in accuracy [3][12]. Group 2: Clinical Implications - The non-invasive model has the potential to improve clinical management of CKD, especially for patients who are unsuitable for traditional kidney biopsies [3][16]. - KIDS provides objective pathological and prognostic predictions, which could enhance kidney care quality and reduce the incidence of end-stage renal disease, particularly in underdeveloped regions [16]. Group 3: Background and Challenges - CKD affects approximately 850 million people globally and poses significant health risks, with kidney biopsy being the gold standard for diagnosis but often limited by various factors [6][7]. - Complications from kidney biopsies, such as bleeding, occur in 11% of cases, highlighting the need for safer diagnostic alternatives [6].