Patients allocated to the COT group are told to expose themselves twice daily to four odors (phenyl ethyl alcohol: rose, menthol: mint, citronellal: lemon, and eugenol: cloves)，which are placed in four brown glass jars (total volume 50 mL) with one of the four odors in each (1 mL each, soaked in cotton pads to prevent spilling). All jars are labeled with the odor name.

Olfactory training includes exposure to odorants twice per day for 5 minutes. Every olfactory session included rotated exposure to each odorant for 10 seconds each, with time intervals of 10 seconds between odors. The resting order of the four odors is mint-rose-lemon-clove.

Patients were advised to sniff the odors in the morning before breakfast and in the evening before bedtime.

The COT group will differ solely with regard to the ingredients contained in the training jars given to the participants. Whereas in the COT group, the training jars will contain the active odors previously selected, in the placebo group, the training jars will be filled with a placebo odorless solvent (propylene glycol). To increase the overall compliance and to ascertain that the placebo (odorless) group remains blinded to the treatment allocation, all patients are told that whereas the odors used during the therapy tend to be less strong, and may be insufficiently intense to produce an odor sensation (i.e., subliminal odors), they are intense enough to be perceived by the brain and may influence mental processes and behavior. The training will last for 12 weeks. After the study completion, all participants will be individually contacted and their group allocations will be revealed. All patients will be offered the possibility of performing or continuing a training with odors.

Conventional olfactory training was designed by Hummel et al. They performed OT using four odors (phenyl ethyl alcohol, rose; eucalyptol, eucalyptus; citronella, lemon; and eugenol, cloves), which are representative of four odor categories on the odor prism proposed by Henning.

This olfactory test consists of three subtests that respectively measure odor threshold (T), odor discrimination (D), and odor identification (I). Each subtest has a maximum score of 16, and the sum of the scores from the three subtests provides the global olfactory score (TDI score; Threshold, Discrimination, Identification) with a maximum of 48 points. Patients with a TDI composite score≥30.5 are defined as normosmic. In contrast, those with a TDI composite score between 16.5 and 30.5 are defined as hyposmic, and patients with a TDI composite score<16.5 are defined as functionally anosmic. Based on previous literature, we have established a 5.5 points improvement in the TDI composite score as a significant change.

OB volumetric evaluations will be performed with a 3-Tesla MRI system with coronal T2-weighted fast spin-echo images that included the anterior and the middle parts of the skull base.

The OB volume will be measured using manual segmentation. The contours of the left and the right OBs will be manually delineated. The proximal limit of the OB will be determined on the basis of the sudden change in the diameter at the beginning of the olfactory tract. After all surfaces are added and multiplied by the slice thickness, the volume in cubic millimeters will be obtained for the right and the left OBs, and the total OB volume will be calculated as the sum of the two values.

Measurements will be performed blindly twice by two examiners. When the difference in volume is > 10%, the examiners will perform a third measurement together.

We have interpreted oval, round, triangular, and bean/boat/banana shapes as normal variants of OB shape.Atrophy will be interpreted as flattened OB.