Table 3.

Excellent clinical outcome AI predictions.

	Outcome +	Outcome −	Total
Prediction +	399	199	598	Sensitivity	74.2%
Prediction −	139	506	645	Specificity	71.8%
Total	538	705	1243	PPV	66.7%
				NPV	78.4%
				Accuracy	72.8%
				AUC ROC	0.808

The table shows the performance evaluation of the AI model predicting the “Excellent Clinical Outcome”—ODI—SF36—COMI Back at 6 months FU. Outcome +: patients with ODI Δ ≥ 12.7/100 and SF-36 PCS (Δ ≥ 6/100) and COMI back (Δ ≥ 2.2/10); Outcome −: patients with at least one of the following conditions ODI Δ < 12.7/100 or SF-36 PCS Δ < 6/100 or COMI back Δ < 2.2/10; Prediction +: model’s predictions of patients with ODI Δ ≥ 12.7/100 and SF-36 PCS Δ ≥ 6/100 and COMI back Δ ≥ 2.2/10; Outcome −: model’s predictions of patients with at least one of the following conditions—ODI Δ < 12.7/100 or SF-36 PCS Δ < 6/100 or COMI back Δ < 2.2/10; PPV: Positive Predictive Values; NPV: Negative Predictive Values