Research Article
Explainable and Personalized Medical Cost Prediction Based on Multitask Learning over Mobile Devices
Table 1
Patient feature categories classification.
| Features | Feature categories | Values |
| Age | 25 - | 0 | 25–40 | 1 | 40–50 | 2 | 50–60 | 3 | 60+ | 4 | T-stage | T0 | 0 | Tis | 1 | Tx | 2 | T1 | 3 | T2 | 4 | T3 | 5 | T4 | 6 | Histologic classification | Invasive carcinoma | 0 | Invasive ductal carcinoma | 1 | Papillary carcinoma | 2 | Infiltrating lobular carcinoma | 3 | Medullary carcinoma | 4 | N-stage | N0 | 0 | N1 | 1 | N2 | 2 | N3 | 3 | Nx | 4 | M-stage | M0 | 0 | M1 | 1 | Mx | 2 | Complication and comorbidity | serious | 0 | General | 1 | — | 2 | HER2 | 0 | 0 | 1+ | 1 | 2+ | 2 | 3+ | 3 |
|
|