SASaud Malfy Alenzy

Saud Malfy Alenzy

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Saud Malfy Alenzy has a total of 9 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • LSTM-based prediction of cognitive impairment and cerebral perfusion in type 2 diabetes using multimodal data for surgical risk assessment

    by Mohamed Yacin Sikkandar, Saud Malfy Alenzy, Sabarunisha Begum S., Rashed Almousa, Jamal Abdulrahman Almosa, Abdulrahman Manajaa Alotaibi, Ahmed Mohammed Aldhfayan, Adel Abdullah Alshehri, Turki Ajab Alotaibi, Yasir Abdullah Alotaibi

    Cognitive impairment (CI) and cerebral perfusion (CP) abnormalities are common in individuals with type 2 diabetes (T2D), contributing to an increased risk of cognitive decline due to vascular complications that reduce cerebral blood flow (CBF). However, most existing studies have examined these issues in isolation and lack integrated predictive models. This study aimed to develop a long short-term memory (LSTM)-based framework to predict both cognitive decline and CP abnormalities in patients with T2D. Data from 50 patients, including key parameters related to diabetic complications, were analyzed. The random forest (RF) method was used for feature selection, ranking variables by importance. The primary prediction targets were the Mini-Mental State Examination (MMSE) total score for cognitive decline and perfusion vascular (whole brain) for CP. LSTM models effectively captured temporal patterns, and performance was evaluated using k-fold cross-validation with mean absolute error (MAE), mean squared error (MSE), and R-squared ([ ]) as metrics. Results demonstrated strong predictive performance: for cognitive decline, the test MAE was 1.35, MSE 4.01, and [ ] 0.35; for CP, the test MAE was 1.05, MSE 2.20, and [ ] 0.44. Integrating RF with LSTM improved prediction accuracy compared to conventional models such as artificial neural networks (ANNs). Clinically, this framework may support early diagnosis, personalized interventions, and risk stratification in patients with diabetes. In particular, accurate prediction of cognitive and perfusion abnormalities could enhance preoperative assessment and surgical planning in diabetic patients undergoing vascular or neurocognitive-related procedures, thereby reducing perioperative risks and improving patient outcomes.