Background: Nicotine and Caffeine are commonly consumed psychostimulant natural alkaloids having profound effects on brain and cardiac bioelectrical signals seen though EEG and ECG respectively and any departure from normal condition can be investigated and predicted using statistical hypothesis tests and machine learning based binary classification. 
 
Methods: The current experiment was performed on adult male Wister rats. 30 animals were randomly divided into 3 categories and administered chronic intraperitoneal doses for 21 days.&nbsp; Category I, the control group, received saline, category II received 0.5 mg/kg Body Weight (BW) nicotine and category III was administered 10 mg/kg BW caffeine. The epidural EEG and non-invasive surface ECG recording were performed on the 21st&nbsp;day under urethane anaesthesia and analyzed on Biopac software. In EEG, total 16 parameters-4 each from delta, theta, alpha and beta bands-maximum frequency, maximum amplitude, mean amplitude and area under curve, were extracted. For ECG, the amplitude of P, R, T waves, heart rate and duration of PR, QRS, QT and RR were selected. Separate Control vs Nicotine and Control vs Caffeine comparisons were made using unpaired t test-based hypothesis testing and support vector machine (SVM)-based automated binary classification. 
 
Result: The t test analysis for both Nicotine and Caffeine groups showed that in EEG, the delta wave maximum frequency and the beta wave band maximum amplitude was decreased (P&lt;0.01). It also showed that in ECG, the amplitude of P, T wave increased along with heart rate (P&lt;0.001). The SVM performed great in binary classification with more than 90% accuracy, precision, recall and F1 score using either EEG or ECG parameters for both Nicotine and Caffeine groups. 
&nbsp;

Neurocardiac Effects of Nicotine and Caffeine in Wistar Rats: An EEG-ECG Study Integrating Inferential Statistics and Machine Learning for Applications in Digital Diagnostics