Establishment of a corneal ulcer prognostic model based on machine learning.

Corneal infection is a major public health concern worldwide and the most common cause of unilateral corneal blindness. Toxic effects of different microorganisms, such as bacteria and fungi, worsen keratitis leading to corneal perforation even with optimal drug treatment. The cornea forms the main refractive surface of the eye. Diseases affecting the cornea can cause severe visual impairment. Therefore, it is crucial to analyze the risk of corneal perforation and visual impairment in corneal ulcer patients for making early treatment strategies. The modeling of a fully automated prognostic model system was performed in two parts. In the first part, the dataset contained 4973 slit lamp images of corneal ulcer patients in three centers. A deep learning model was developed and tested for segmenting and classifying five lesions (corneal ulcer, corneal scar, hypopyon, corneal descementocele, and corneal neovascularization) in the eyes of corneal ulcer patients. Further, hierarchical quantification was carried out based on policy rules. In the second part, the dataset included clinical data (name, gender, age, best corrected visual acuity, and type of corneal ulcer) of 240 patients with corneal ulcers and respective 1010 slit lamp images under two light sources (natural light and cobalt blue light). The slit lamp images were then quantified hierarchically according to the policy rules developed in the first part of the modeling. Combining the above clinical data, the features were used to build the final prognostic model system for corneal ulcer perforation outcome and visual impairment using machine learning algorithms such as XGBoost, LightGBM. The ROC curve area (AUC value) evaluated the model's performance. For segmentation of the five lesions, the accuracy rates of hypopyon, descemetocele, corneal ulcer under blue light, and corneal neovascularization were 96.86, 91.64, 90.51, and 93.97, respectively. For the corneal scar lesion classification, the accuracy rate of the final model was 69.76. The XGBoost model performed the best in predicting the 1-month prognosis of patients, with an AUC of 0.81 (95% CI 0.63-1.00) for ulcer perforation and an AUC of 0.77 (95% CI 0.63-0.91) for visual impairment. In predicting the 3-month prognosis of patients, the XGBoost model received the best AUC of 0.97 (95% CI 0.92-1.00) for ulcer perforation, while the LightGBM model achieved the best performance with an AUC of 0.98 (95% CI 0.94-1.00) for visual impairment.

Using intramyocardial electrograms combined with other noninvasive methods for monitoring acute rejection following human heart transplantation.

BACKGROUND: Acute allograft rejection in heart transplantation remains as one of the major complications. Obligatory graft surveillance is still achieved with the invasive and expensive endomyocardial biopsy (EMB). Our study aimed to study the use of intramyocardial electrograms combined with other noninvasive methods for the monitoring of acute rejection after human heart transplantation. METHODS: Permanent pacemakers were implanted in 58 patients undergoing heart transplantations. Intramyocardial electrograms (IMEG) were recorded periodically and the results were compared with those from EMBs. The R wave amplitude of the IMEG was used as the index value, the average R wave amplitude at the third week following transplantation was considered as the baseline, and a reduction of > 20% compared with the baseline was regarded as a positive result. EMB was performed in cases of positive IMEG results and also at other times. Other noninvasive methods were used to help the diagnosis. Acute rejection (AR) was defined as International Society of Heart-Lung Transplantation grade IIIA or higher. RESULTS: We obtained 1231 IMEG records and 127 EMBs. Of the total 127 EMBs, 53 were positive, in which there were 42 IMEG positive results and 11 negative, while in the rest 74 negative EMBs, there were 9 IMEG positive results and 65 negative. The sensitivity of IMEG for the diagnosis of AR was 79.2%, and the specificity was 87.8%. The positive predictive value was 82.4% and the negative predictive value was 85.5%. Of the total of 1231 IMEG records, 51 were positive and 1180 were negative. Excluding 11 proved by EMB to be false negative, if the other 1169 were considered as no evidence of rejection, through the other noninvasive methods, AR diagnosed by this noninvasive monitoring strategy, the sensitivity was 79.2%, and the specificity was 99.2%. The positive predictive value was 82.4% and the negative predictive value was 99.1%. CONCLUSIONS: IMEG can be used as a noninvasive method for monitoring AR following heart transplantation. It is a continuous, safe and inexpensive method, and could reduce the need for EMB combined with other noninvasive methods, without reducing the detection of rejection.

Intraoperative modified Cox mini-maze procedure for long-standing persistent atrial fibrillation.

BACKGROUND: As the most effective method of curing atrial fibrillation, the classic Maze III procedure eliminates atrial fibrillation in more than 90% of patients. In spite of high efficacy, the complexity and complications associated with this procedure prevents widespread application. In this research, we report our experience with a modified Cox mini-Maze procedure using radiofrequency energy for treating long-standing persistent atrial fibrillation during open-heart surgery. METHODS: A total of 45 patients with long-standing persistent atrial fibrillation who underwent open-heart surgery received the treatment. This modified Cox mini-Maze procedure was performed using unipolar and bipolar radiofrequency systems. Some lesions were done with unipolar pen both epicardially and endocardially to improve the transmurality of ablation. RESULTS: Hospital mortality was 0%. The follow-up time ranged from 6 to 26 months (mean, 16.6). Overall freedom from atrial fibrillation was 86.7% at latest follow-up after surgery. When analyzed by the Kaplan-Meier method, freedom from atrial fibrillation was 94% (6 months), 87% (12 months), and 82% (24 months). Compared with a randomized Cox Maze III control group, the results of freedom from atrial fibrillation do not show significant differences between the two groups at interval contacts. The survival rate was 100% (45 of 45), as was freedom from stroke. No patients required implantation of a permanent pacemaker. Recovery of atrial contractility occurred in 94.6% of patients (35 of 37) during the latest follow-up. CONCLUSIONS: According to our experience, the modified Cox mini-Maze procedure allows rapid ablation application and offers an optimistic outcome for the recovery of sinus rhythm in patients with long-standing persistent atrial fibrillation who undergo open-heart surgery.